def get_match_first(lits, parseAction=None):
el = MatchFirst(NoMatch())
for lit in lits:
el = el.__ior__(lit)
if parseAction:
el.setParseAction(parseAction)
return el
def get_match_first(lits, parseAction=None):
el = MatchFirst(NoMatch())
for lit in lits:
el = el.__ior__(lit)
if parseAction:
el.setParseAction(parseAction)
return el
def define_simple_literals(literal_list, parseAction=None):
l = MatchFirst([CaselessKeyword(k) for k in literal_list])
if parseAction:
l = l.setParseAction(parseAction)
return l
def define_simple_literals(literal_list, parseAction=None):
l = MatchFirst([CaselessKeyword(k) for k in literal_list])
if parseAction:
l = l.setParseAction(parseAction)
return l
def _get_requirements_build_gradle(path: str) -> list:
"""
Get list of requirements from Maven project.
Files supported are build.gradle
:param path: Project path
"""
reqs = []
for file_path in full_paths_in_dir(path):
if not file_path.endswith('build.gradle'):
continue
with open(file_path, encoding='latin-1') as file_fd:
file_content = file_fd.read()
string = MatchFirst([quotedString('"'), quotedString("'")])
string.setParseAction(lambda x: [x[0][1:-1]])
grammars: list = [
Suppress(Keyword('compile') + Optional('(')) +
string.copy()('package'),
Suppress(Keyword('compile') + Optional('(')) +
Suppress(Keyword('group') + ':') + string.copy()('group') +
Suppress(',') + Suppress(Keyword('name') + ':') +
string.copy()('name') + Suppress(',') +
Suppress(Keyword('version') + ':') + string.copy()('version'),
]
for grammar in grammars:
for tokens, _, _ in grammar.scanString(file_content):
matches = tokens.asDict()
if 'package' in matches:
if ':' in matches['package']:
name, version = matches['package'].rsplit(':', 1)
else:
name, version = matches['package'], None
reqs.append((file_path, name, version))
else:
reqs.append(
(file_path, f"{matches['group']}:{matches['name']}",
matches['version']))
reqs.append(
(file_path, matches['group'], matches['version']))
return reqs
def __init__(self):
self.json_query = {'query':{}, 'and': [], 'or': []}
self.tokens = None
#--------------------------------------------------------------------------------------
# <integer> ::= 0-9
# <double> ::= 0-9 ('.' 0-9)
# <number> ::= <integer> | <double>
#--------------------------------------------------------------------------------------
integer = Regex(r'-?[0-9]+') # Word matches space for some reason
double = Regex(r'-?[0-9]+.?[0-9]*')
number = double | integer
#--------------------------------------------------------------------------------------
# <python-string> ::= (String surrounded by double-quotes)
# <wildcard-string> ::= <python-string>
# <limited-string> ::= '"' a..z A..Z 9..9 _ . '"' (alpha nums and ._ surrounded by double quotes)
# <field> ::= <limited-string> | "*"
# <coords> ::= "LAT" <number> "LON" <number>
# <units> ::= ('km' | 'mi' | 'nm')
# <distance> ::= REGEX(([0-9]*\.?[0-9]*)(km|mi|nm)?)
#--------------------------------------------------------------------------------------
python_string = quotedString.setParseAction(removeQuotes)
wildcard_string = python_string
limited_string = Regex(r'("(?:[a-zA-Z0-9_\.])*"|\'(?:[a-zA-Z0-9_\.]*)\')').setParseAction(removeQuotes)
field = limited_string ^ CaselessLiteral('"*"').setParseAction(removeQuotes)
coords = CaselessLiteral("LAT") + number + CaselessLiteral("LON") + number
units = CaselessLiteral('km') | CaselessLiteral('mi')
distance = number + units
distance.setParseAction( lambda x : self.frame.update({'dist' : float(x[0]), 'units' : x[1]}))
#--------------------------------------------------------------------------------------
# Date
#--------------------------------------------------------------------------------------
date = python_string
#--------------------------------------------------------------------------------------
# <query-filter> ::= "FILTER" <python-string>
# <index-name> ::= <python-string>
# <resource-id> ::= '"' a..z A..Z 0..9 $ _ -'"' (alpha nums surrounded by double quotes)
# <collection-id> ::= <resource-id>
#--------------------------------------------------------------------------------------
query_filter = CaselessLiteral("FILTER") + python_string
# Add the filter to the frame object
query_filter.setParseAction(lambda x : self.frame.update({'filter' : x[1]}))
index_name = MatchFirst(python_string)
# Add the index to the frame object
index_name.setParseAction(lambda x : self.frame.update({'index' : x[0]}))
resource_id = Regex(r'("(?:[a-zA-Z0-9\$_-])*"|\'(?:[a-zA-Z0-9\$_-]*)\')').setParseAction(removeQuotes)
collection_id = resource_id
#--------------------------------------------------------------------------------------
# <from-statement> ::= "FROM" <number>
# <to-statement> ::= "TO" <number>
#--------------------------------------------------------------------------------------
from_statement = CaselessLiteral("FROM") + number
from_statement.setParseAction(lambda x : self.frame.update({'from' : x[1]}))
to_statement = CaselessLiteral("TO") + number
to_statement.setParseAction(lambda x : self.frame.update({'to' : x[1]}))
#--------------------------------------------------------------------------------------
# <date-from-statement> ::= "FROM" <date>
# <date-to-statement> ::= "TO" <date>
#--------------------------------------------------------------------------------------
date_from_statement = CaselessLiteral("FROM") + date
date_from_statement.setParseAction(lambda x : self.frame.update({'from' : x[1]}))
date_to_statement = CaselessLiteral("TO") + date
date_to_statement.setParseAction(lambda x : self.frame.update({'to' : x[1]}))
#--------------------------------------------------------------------------------------
# <time-query> ::= "TIME FROM" <date> "TO" <date>
#--------------------------------------------------------------------------------------
time_query = CaselessLiteral("TIME") + Optional(date_from_statement) + Optional(date_to_statement)
time_query.setParseAction(lambda x : self.time_frame())
# time.mktime(dateutil.parser.parse(x[2])), 'to':time.mktime(dateutil.parser.parse(x[4]))}}))
#--------------------------------------------------------------------------------------
# <time-bounds> ::= "TIMEBOUNDS" <from-statement> <to-statement>
#--------------------------------------------------------------------------------------
time_bounds = CaselessLiteral("TIMEBOUNDS") + date_from_statement + date_to_statement
time_bounds.setParseAction(lambda x : self.time_bounds_frame())
#--------------------------------------------------------------------------------------
# <vertical-bounds> ::= "VERTICAL" <from-statement> <to-statement>
#--------------------------------------------------------------------------------------
vertical_bounds = CaselessLiteral("VERTICAL") + from_statement + to_statement
vertical_bounds.setParseAction(lambda x : self.vertical_bounds_frame())
#--------------------------------------------------------------------------------------
# <range-query> ::= "VALUES" [<from-statement>] [<to-statement>]
#--------------------------------------------------------------------------------------
range_query = CaselessLiteral("VALUES") + Optional(from_statement) + Optional(to_statement)
# Add the range to the frame object
range_query.setParseAction(lambda x : self.range_frame())
#--------------------------------------------------------------------------------------
# <geo-distance> ::= "DISTANCE" <distance> "FROM" <coords>
# <geo-bbox> ::= "BOX" "TOP-LEFT" <coords> "BOTTOM-RIGHT" <coords>
#--------------------------------------------------------------------------------------
geo_distance = CaselessLiteral("DISTANCE") + distance + CaselessLiteral("FROM") + coords
geo_distance.setParseAction(lambda x : self.frame.update({'lat': float(x[5]), 'lon':float(x[7])}))
geo_bbox = CaselessLiteral("BOX") + CaselessLiteral("TOP-LEFT") + coords + CaselessLiteral("BOTTOM-RIGHT") + coords
geo_bbox.setParseAction(lambda x : self.frame.update({'top_left':[float(x[5]),float(x[3])], 'bottom_right':[float(x[10]),float(x[8])]}))
#--------------------------------------------------------------------------------------
# <field-query> ::= <wildcard-string>
# <term-query> ::= "IS" <field-query>
# <fuzzy-query> ::= "LIKE" <field-query>
# <match-query> ::= "MATCH" <field-query>
# <geo-query> ::= "GEO" ( <geo-distance> | <geo-bbox> )
#--------------------------------------------------------------------------------------
field_query = wildcard_string
term_query = CaselessLiteral("IS") + field_query
term_query.setParseAction(lambda x : self.frame.update({'value':x[1]}))
geo_query = CaselessLiteral("GEO") + ( geo_distance | geo_bbox )
fuzzy_query = CaselessLiteral("LIKE") + field_query
fuzzy_query.setParseAction(lambda x : self.frame.update({'fuzzy':x[1]}))
match_query = CaselessLiteral("MATCH") + field_query
match_query.setParseAction(lambda x : self.frame.update({'match':x[1]}))
#--------------------------------------------------------------------------------------
# <limit-parameter> ::= "LIMIT" <integer>
# <depth-parameter> ::= "DEPTH" <integer>
# <order-parameter> ::= "ORDER" "BY" <limited-string>
# <offset-parameter> ::= "SKIP" <integer>
# <query-parameter> ::= <order-paramater> | <limit-parameter>
#--------------------------------------------------------------------------------------
limit_parameter = CaselessLiteral("LIMIT") + integer
limit_parameter.setParseAction(lambda x: self.json_query.update({'limit' : int(x[1])}))
depth_parameter = CaselessLiteral("DEPTH") + integer
depth_parameter.setParseAction(lambda x: self.frame.update({'depth' : int(x[1])}))
order_parameter = CaselessLiteral("ORDER") + CaselessLiteral("BY") + limited_string
order_parameter.setParseAction(lambda x: self.json_query.update({'order' : {x[2] : 'asc'}}))
offset_parameter = CaselessLiteral("SKIP") + integer
offset_parameter.setParseAction(lambda x : self.json_query.update({'skip' : int(x[1])}))
query_parameter = limit_parameter | order_parameter | offset_parameter
#--------------------------------------------------------------------------------------
# <search-query> ::= "SEARCH" <field> (<range-query> | <term-query> | <fuzzy-query> | <match-query> | <time-query> | <time-bounds> | <vertical-bounds> | <geo-query>) "FROM" <index-name> [<query-parameter>]*
# <collection-query> ::= "IN <collection-id>"
# <association-query> ::= "BELONGS TO" <resource-id> [ <depth-parameter> ]
# <owner-query> ::= "HAS" <resource-id> [ <depth-parameter> ]
# <query> ::= <search-query> | <association-query> | <collection-query> | <owner-query>
#--------------------------------------------------------------------------------------
search_query = CaselessLiteral("SEARCH") + field + (range_query | term_query | fuzzy_query | match_query | vertical_bounds | time_bounds | time_query | geo_query) + CaselessLiteral("FROM") + index_name
# Add the field to the frame object
search_query.setParseAction(lambda x : self.frame.update({'field' : x[1]}))
collection_query = CaselessLiteral("IN") + collection_id
collection_query.setParseAction(lambda x : self.frame.update({'collection': x[1]}))
association_query = CaselessLiteral("BELONGS") + CaselessLiteral("TO") + resource_id + Optional(depth_parameter)
# Add the association to the frame object
association_query.setParseAction(lambda x : self.frame.update({'association':x[2]}))
owner_query = CaselessLiteral("HAS") + resource_id + Optional(depth_parameter)
owner_query.setParseAction(lambda x : self.frame.update({'owner':x[1]}))
query = search_query | association_query | collection_query | owner_query
#--------------------------------------------------------------------------------------
# <primary-query> ::= <query> [<query-filter>]
# <atom> ::= <query>
# <intersection> ::= "AND" <atom>
# <union> ::= "OR" <atom>
# <sentence> ::= <primary-query> [<intersection>]* [<union>]*
#--------------------------------------------------------------------------------------
primary_query = query + Optional(query_filter)
# Set the primary query on the json_query to the frame and clear the frame
primary_query.setParseAction(lambda x : self.push_frame())
atom = query
intersection = CaselessLiteral("AND") + atom
# Add an AND operation to the json_query and clear the frame
intersection.setParseAction(lambda x : self.and_frame())
union = CaselessLiteral("OR") + atom
# Add an OR operation to the json_query and clear the frame
union.setParseAction(lambda x : self.or_frame())
self.sentence = primary_query + (intersection ^ union)*(0,None) + query_parameter*(0,None)
class BELParser(BaseParser):
"""Build a parser backed by a given dictionary of namespaces."""
def __init__(
self,
graph,
namespace_to_term: Optional[Mapping[str, Mapping[str, str]]] = None,
namespace_to_pattern: Optional[Mapping[str, Pattern]] = None,
annotation_to_term: Optional[Mapping[str, Set[str]]] = None,
annotation_to_pattern: Optional[Mapping[str, Pattern]] = None,
annotation_to_local: Optional[Mapping[str, Set[str]]] = None,
allow_naked_names: bool = False,
allow_nested: bool = False,
disallow_unqualified_translocations: bool = False,
citation_clearing: bool = True,
skip_validation: bool = False,
autostreamline: bool = True,
required_annotations: Optional[List[str]] = None,
) -> None:
"""Build a BEL parser.
:param pybel.BELGraph graph: The BEL Graph to use to store the network
:param namespace_to_term: A dictionary of {namespace: {name: encoding}}. Delegated to
:class:`pybel.parser.parse_identifier.IdentifierParser`
:param namespace_to_pattern: A dictionary of {namespace: regular expression strings}. Delegated to
:class:`pybel.parser.parse_identifier.IdentifierParser`
:param annotation_to_term: A dictionary of {annotation: set of values}. Delegated to
:class:`pybel.parser.ControlParser`
:param annotation_to_pattern: A dictionary of {annotation: regular expression strings}. Delegated to
:class:`pybel.parser.ControlParser`
:param annotation_to_local: A dictionary of {annotation: set of values}. Delegated to
:class:`pybel.parser.ControlParser`
:param allow_naked_names: If true, turn off naked namespace failures. Delegated to
:class:`pybel.parser.parse_identifier.IdentifierParser`
:param allow_nested: If true, turn off nested statement failures. Delegated to
:class:`pybel.parser.parse_identifier.IdentifierParser`
:param disallow_unqualified_translocations: If true, allow translocations without TO and FROM clauses.
:param citation_clearing: Should :code:`SET Citation` statements clear evidence and all annotations?
Delegated to :class:`pybel.parser.ControlParser`
:param autostreamline: Should the parser be streamlined on instantiation?
:param required_annotations: Optional list of required annotations
"""
self.graph = graph
self.allow_nested = allow_nested
self.disallow_unqualified_translocations = disallow_unqualified_translocations
if skip_validation:
self.control_parser = ControlParser(
citation_clearing=citation_clearing,
required_annotations=required_annotations,
)
self.identifier_parser = IdentifierParser(
allow_naked_names=allow_naked_names, )
else:
self.control_parser = ControlParser(
annotation_to_term=annotation_to_term,
annotation_to_pattern=annotation_to_pattern,
annotation_to_local=annotation_to_local,
citation_clearing=citation_clearing,
required_annotations=required_annotations,
)
self.identifier_parser = IdentifierParser(
allow_naked_names=allow_naked_names,
namespace_to_term=namespace_to_term,
namespace_to_pattern=namespace_to_pattern,
)
self.control_parser.get_line_number = self.get_line_number
self.identifier_parser.get_line_number = self.get_line_number
identifier = Group(self.identifier_parser.language)(IDENTIFIER)
ungrouped_identifier = self.identifier_parser.language
# 2.2 Abundance Modifier Functions
#: `2.2.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_protein_modifications>`_
self.pmod = get_protein_modification_language(
self.identifier_parser.identifier_qualified)
#: `2.2.4 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_cellular_location>`_
self.location = get_location_language(self.identifier_parser.language)
opt_location = pyparsing.Optional(WCW + self.location)
#: PyBEL BEL Specification variant
self.gmod = get_gene_modification_language(
self.identifier_parser.identifier_qualified)
# 2.6 Other Functions
#: `2.6.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_fusion_fus>`_
self.fusion = get_fusion_language(self.identifier_parser.language)
# 2.1 Abundance Functions
#: `2.1.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XcomplexA>`_
self.general_abundance = general_abundance_tags + nest(
ungrouped_identifier + opt_location)
self.gene_modified = ungrouped_identifier + pyparsing.Optional(
WCW + delimitedList(Group(variant | gsub | self.gmod))(VARIANTS))
self.gene_fusion = Group(self.fusion)(FUSION)
self.gene_fusion_legacy = Group(
get_legacy_fusion_langauge(identifier, 'c'))(FUSION)
#: `2.1.4 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XgeneA>`_
self.gene = gene_tag + nest(
MatchFirst([
self.gene_fusion, self.gene_fusion_legacy, self.gene_modified
]) + opt_location)
self.mirna_modified = ungrouped_identifier + pyparsing.Optional(
WCW + delimitedList(Group(variant))(VARIANTS)) + opt_location
#: `2.1.5 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XmicroRNAA>`_
self.mirna = mirna_tag + nest(self.mirna_modified)
self.protein_modified = ungrouped_identifier + pyparsing.Optional(
WCW + delimitedList(
Group(MatchFirst([self.pmod, variant, fragment, psub, trunc])))
(VARIANTS))
self.protein_fusion = Group(self.fusion)(FUSION)
self.protein_fusion_legacy = Group(
get_legacy_fusion_langauge(identifier, 'p'))(FUSION)
#: `2.1.6 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XproteinA>`_
self.protein = protein_tag + nest(
MatchFirst([
self.protein_fusion,
self.protein_fusion_legacy,
self.protein_modified,
]) + opt_location)
self.rna_modified = ungrouped_identifier + pyparsing.Optional(
WCW + delimitedList(Group(variant))(VARIANTS))
self.rna_fusion = Group(self.fusion)(FUSION)
self.rna_fusion_legacy = Group(
get_legacy_fusion_langauge(identifier, 'r'))(FUSION)
#: `2.1.7 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XrnaA>`_
self.rna = rna_tag + nest(
MatchFirst([
self.rna_fusion,
self.rna_fusion_legacy,
self.rna_modified,
]) + opt_location)
self.single_abundance = MatchFirst([
self.general_abundance, self.gene, self.mirna, self.protein,
self.rna
])
#: `2.1.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XcomplexA>`_
self.complex_singleton = complex_tag + nest(ungrouped_identifier +
opt_location)
self.complex_list = complex_tag + nest(
delimitedList(Group(self.single_abundance
| self.complex_singleton))(MEMBERS) +
opt_location)
self.complex_abundances = self.complex_list | self.complex_singleton
# Definition of all simple abundances that can be used in a composite abundance
self.simple_abundance = self.complex_abundances | self.single_abundance
self.simple_abundance.setParseAction(self.check_function_semantics)
#: `2.1.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XcompositeA>`_
self.composite_abundance = composite_abundance_tag + nest(
delimitedList(Group(self.simple_abundance))(MEMBERS) +
opt_location)
self.abundance = self.simple_abundance | self.composite_abundance
# 2.4 Process Modifier Function
# backwards compatibility with BEL v1.0
molecular_activity_default = oneOf(list(
language.activity_labels)).setParseAction(
handle_molecular_activity_default)
#: `2.4.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XmolecularA>`_
self.molecular_activity = molecular_activity_tags + nest(
molecular_activity_default | self.identifier_parser.language)
# 2.3 Process Functions
#: `2.3.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_biologicalprocess_bp>`_
self.biological_process = biological_process_tag + nest(
ungrouped_identifier)
#: `2.3.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_pathology_path>`_
self.pathology = pathology_tag + nest(ungrouped_identifier)
self.bp_path = self.biological_process | self.pathology
self.bp_path.setParseAction(self.check_function_semantics)
self.activity_standard = activity_tag + nest(
Group(self.simple_abundance)(TARGET) +
pyparsing.Optional(WCW + Group(self.molecular_activity)(EFFECT)))
activity_legacy_tags = oneOf(language.activities)(MODIFIER)
self.activity_legacy = activity_legacy_tags + nest(
Group(self.simple_abundance)(TARGET))
self.activity_legacy.setParseAction(handle_activity_legacy)
#: `2.3.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#Xactivity>`_
self.activity = self.activity_standard | self.activity_legacy
self.process = self.bp_path | self.activity
# 2.5 Transformation Functions
from_loc = Suppress(FROM_LOC) + nest(identifier(FROM_LOC))
to_loc = Suppress(TO_LOC) + nest(identifier(TO_LOC))
self.cell_secretion = cell_secretion_tag + nest(
Group(self.simple_abundance)(TARGET))
self.cell_surface_expression = cell_surface_expression_tag + nest(
Group(self.simple_abundance)(TARGET))
self.translocation_standard = nest(
Group(self.simple_abundance)(TARGET) + WCW +
Group(from_loc + WCW + to_loc)(EFFECT))
self.translocation_legacy = nest(
Group(self.simple_abundance)(TARGET) + WCW +
Group(identifier(FROM_LOC) + WCW + identifier(TO_LOC))(EFFECT))
self.translocation_legacy.addParseAction(handle_legacy_tloc)
self.translocation_unqualified = nest(
Group(self.simple_abundance)(TARGET))
if self.disallow_unqualified_translocations:
self.translocation_unqualified.setParseAction(
self.handle_translocation_illegal)
#: `2.5.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_translocations>`_
self.translocation = translocation_tag + MatchFirst([
self.translocation_unqualified, self.translocation_standard,
self.translocation_legacy
])
#: `2.5.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_degradation_deg>`_
self.degradation = degradation_tags + nest(
Group(self.simple_abundance)(TARGET))
#: `2.5.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_reaction_rxn>`_
self.reactants = Suppress(REACTANTS) + nest(
delimitedList(Group(self.simple_abundance)))
self.products = Suppress(PRODUCTS) + nest(
delimitedList(Group(self.simple_abundance)))
self.reaction = reaction_tags + nest(
Group(self.reactants)(REACTANTS),
Group(self.products)(PRODUCTS))
self.transformation = MatchFirst([
self.cell_secretion, self.cell_surface_expression,
self.translocation, self.degradation, self.reaction
])
# 3 BEL Relationships
self.bel_term = MatchFirst(
[self.transformation, self.process, self.abundance]).streamline()
self.bel_to_bel_relations = [
association_tag,
increases_tag,
decreases_tag,
positive_correlation_tag,
negative_correlation_tag,
causes_no_change_tag,
orthologous_tag,
is_a_tag,
equivalent_tag,
partof_tag,
directly_increases_tag,
directly_decreases_tag,
analogous_tag,
regulates_tag,
]
self.bel_to_bel = triple(self.bel_term,
MatchFirst(self.bel_to_bel_relations),
self.bel_term)
# Mixed Relationships
#: `3.1.5 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_ratelimitingstepof>`_
self.rate_limit = triple(
MatchFirst(
[self.biological_process, self.activity, self.transformation]),
rate_limit_tag, self.biological_process)
#: `3.4.6 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_subprocessof>`_
self.subprocess_of = triple(
MatchFirst([self.process, self.activity, self.transformation]),
subprocess_of_tag, self.process)
#: `3.3.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_transcribedto>`_
self.transcribed = triple(self.gene, transcribed_tag, self.rna)
#: `3.3.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_translatedto>`_
self.translated = triple(self.rna, translated_tag, self.protein)
#: `3.4.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_hasmember>`_
self.has_member = triple(self.abundance, has_member_tag,
self.abundance)
#: `3.4.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_hasmembers>`_
self.abundance_list = Suppress('list') + nest(
delimitedList(Group(self.abundance)))
self.has_members = triple(self.abundance, has_members_tag,
self.abundance_list)
self.has_members.setParseAction(self.handle_has_members)
self.has_components = triple(self.abundance, has_components_tag,
self.abundance_list)
self.has_components.setParseAction(self.handle_has_components)
self.has_list = self.has_members | self.has_components
# `3.4.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_hascomponent>`_
self.has_component = triple(
self.complex_abundances | self.composite_abundance,
has_component_tag, self.abundance)
self.biomarker = triple(self.bel_term, biomarker_tags, self.process)
self.has_variant_relation = triple(self.abundance, has_variant_tags,
self.abundance)
self.part_of_reaction = triple(self.reaction, part_of_reaction_tags,
self.abundance)
self.relation = MatchFirst([
self.bel_to_bel,
# self.has_member,
# self.has_component,
self.subprocess_of,
self.rate_limit,
self.biomarker,
self.transcribed,
self.translated,
# self.has_variant_relation,
# self.part_of_reaction,
])
self.relation.setParseAction(self._handle_relation_harness)
self.unqualified_relation = MatchFirst([
self.has_member, self.has_component, self.has_variant_relation,
self.part_of_reaction
])
self.unqualified_relation.setParseAction(
self.handle_unqualified_relation)
#: 3.1 Causal Relationships - nested. Not enabled by default.
causal_relation_tags = MatchFirst([
increases_tag, decreases_tag, directly_decreases_tag,
directly_increases_tag
])
self.nested_causal_relationship = triple(
self.bel_term, causal_relation_tags,
nest(triple(self.bel_term, causal_relation_tags, self.bel_term)))
self.nested_causal_relationship.setParseAction(
self.handle_nested_relation)
self.label_relationship = And([
Group(self.bel_term)(SUBJECT),
Suppress('labeled'),
quote(OBJECT)
])
self.label_relationship.setParseAction(self.handle_label_relation)
# has_members is handled differently from all other relations becuase it gets distrinbuted
self.relation = MatchFirst([
self.has_list,
self.nested_causal_relationship,
self.relation,
self.unqualified_relation,
self.label_relationship,
])
self.singleton_term = (self.bel_term + StringEnd()).setParseAction(
self.handle_term)
self.statement = self.relation | self.singleton_term
self.language = self.control_parser.language | self.statement
self.language.setName('BEL')
super(BELParser, self).__init__(self.language,
streamline=autostreamline)
@property
def _namespace_dict(self) -> Mapping[str, Mapping[str, str]]:
"""Get the dictionary of {namespace: {name: encoding}} stored in the internal identifier parser."""
return self.identifier_parser.namespace_to_terms
@property
def _allow_naked_names(self) -> bool:
"""Return if naked names should be parsed (``True``), or if errors should be thrown (``False``)."""
return self.identifier_parser.allow_naked_names
def get_annotations(self) -> Dict:
"""Get the current annotations in this parser."""
return self.control_parser.get_annotations()
def clear(self):
"""Clear the graph and all control parser data (current citation, annotations, and statement group)."""
self.graph.clear()
self.control_parser.clear()
def handle_nested_relation(self, line: str, position: int,
tokens: ParseResults):
"""Handle nested statements.
If :code:`allow_nested` is False, raises a ``NestedRelationWarning``.
:raises: NestedRelationWarning
"""
if not self.allow_nested:
raise NestedRelationWarning(self.get_line_number(), line, position)
self._handle_relation_harness(
line, position, {
SUBJECT: tokens[SUBJECT],
RELATION: tokens[RELATION],
OBJECT: tokens[OBJECT][SUBJECT],
})
self._handle_relation_harness(
line, position, {
SUBJECT: tokens[OBJECT][SUBJECT],
RELATION: tokens[OBJECT][RELATION],
OBJECT: tokens[OBJECT][OBJECT],
})
return tokens
def check_function_semantics(self, line: str, position: int,
tokens: ParseResults) -> ParseResults:
"""Raise an exception if the function used on the tokens is wrong.
:raises: InvalidFunctionSemantic
"""
if not self._namespace_dict or NAMESPACE not in tokens:
return tokens
namespace, name = tokens[NAMESPACE], tokens[NAME]
if namespace in self.identifier_parser.namespace_to_pattern:
return tokens
if self._allow_naked_names and tokens[
NAMESPACE] == DIRTY: # Don't check dirty names in lenient mode
return tokens
valid_functions = set(
itt.chain.from_iterable(
belns_encodings.get(k, set())
for k in self._namespace_dict[namespace][name]))
if not valid_functions:
raise InvalidEntity(self.get_line_number(), line, position,
namespace, name)
if tokens[FUNCTION] not in valid_functions:
raise InvalidFunctionSemantic(self.get_line_number(), line,
position, tokens[FUNCTION],
namespace, name, valid_functions)
return tokens
def handle_term(self, _, __, tokens: ParseResults) -> ParseResults:
"""Handle BEL terms (the subject and object of BEL relations)."""
self.ensure_node(tokens)
return tokens
def _handle_list_helper(self, tokens: ParseResults,
relation: str) -> ParseResults:
"""Provide the functionality for :meth:`handle_has_members` and :meth:`handle_has_components`."""
parent_node_dsl = self.ensure_node(tokens[0])
for child_tokens in tokens[2]:
child_node_dsl = self.ensure_node(child_tokens)
self.graph.add_unqualified_edge(parent_node_dsl, child_node_dsl,
relation)
return tokens
def handle_has_members(self, _, __, tokens: ParseResults) -> ParseResults:
"""Handle list relations like ``p(X) hasMembers list(p(Y), p(Z), ...)``."""
return self._handle_list_helper(tokens, HAS_MEMBER)
def handle_has_components(self, _, __,
tokens: ParseResults) -> ParseResults:
"""Handle list relations like ``p(X) hasComponents list(p(Y), p(Z), ...)``."""
return self._handle_list_helper(tokens, HAS_COMPONENT)
def _add_qualified_edge_helper(self, u, v, relation, annotations,
subject_modifier, object_modifier) -> str:
"""Add a qualified edge from the internal aspects of the parser."""
return self.graph.add_qualified_edge(
u,
v,
relation=relation,
evidence=self.control_parser.evidence,
citation=self.control_parser.citation.copy(),
annotations=annotations,
subject_modifier=subject_modifier,
object_modifier=object_modifier,
**{LINE: self.get_line_number()})
def _add_qualified_edge(self, u, v, relation, annotations,
subject_modifier, object_modifier) -> str:
"""Add an edge, then adds the opposite direction edge if it should."""
sha512 = self._add_qualified_edge_helper(
u,
v,
relation=relation,
annotations=annotations,
subject_modifier=subject_modifier,
object_modifier=object_modifier,
)
if relation in TWO_WAY_RELATIONS:
self._add_qualified_edge_helper(
v,
u,
relation=relation,
annotations=annotations,
object_modifier=subject_modifier,
subject_modifier=object_modifier,
)
return sha512
def _handle_relation(self, tokens: ParseResults) -> str:
"""Handle a relation."""
subject_node_dsl = self.ensure_node(tokens[SUBJECT])
object_node_dsl = self.ensure_node(tokens[OBJECT])
subject_modifier = modifier_po_to_dict(tokens[SUBJECT])
object_modifier = modifier_po_to_dict(tokens[OBJECT])
annotations = {
annotation_name: ({ae: True
for ae in annotation_entry} if isinstance(
annotation_entry, set) else {
annotation_entry: True
})
for annotation_name, annotation_entry in
self.control_parser.annotations.items()
}
return self._add_qualified_edge(
subject_node_dsl,
object_node_dsl,
relation=tokens[RELATION],
annotations=annotations,
subject_modifier=subject_modifier,
object_modifier=object_modifier,
)
def _handle_relation_harness(
self, line: str, position: int,
tokens: Union[ParseResults, Dict]) -> ParseResults:
"""Handle BEL relations based on the policy specified on instantiation.
Note: this can't be changed after instantiation!
"""
if not self.control_parser.citation:
raise MissingCitationException(self.get_line_number(), line,
position)
if not self.control_parser.evidence:
raise MissingSupportWarning(self.get_line_number(), line, position)
missing_required_annotations = self.control_parser.get_missing_required_annotations(
)
if missing_required_annotations:
raise MissingAnnotationWarning(self.get_line_number(), line,
position,
missing_required_annotations)
self._handle_relation(tokens)
return tokens
def handle_unqualified_relation(self, _, __,
tokens: ParseResults) -> ParseResults:
"""Handle unqualified relations."""
subject_node_dsl = self.ensure_node(tokens[SUBJECT])
object_node_dsl = self.ensure_node(tokens[OBJECT])
relation = tokens[RELATION]
self.graph.add_unqualified_edge(subject_node_dsl, object_node_dsl,
relation)
return tokens
def handle_label_relation(self, line: str, position: int,
tokens: ParseResults) -> ParseResults:
"""Handle statements like ``p(X) label "Label for X"``.
:raises: RelabelWarning
"""
subject_node_dsl = self.ensure_node(tokens[SUBJECT])
description = tokens[OBJECT]
if self.graph.has_node_description(subject_node_dsl):
raise RelabelWarning(
line_number=self.get_line_number(),
line=line,
position=position,
node=self.graph.node,
old_label=self.graph.get_node_description(subject_node_dsl),
new_label=description)
self.graph.set_node_description(subject_node_dsl, description)
return tokens
def ensure_node(self, tokens: ParseResults) -> BaseEntity:
"""Turn parsed tokens into canonical node name and makes sure its in the graph."""
if MODIFIER in tokens:
return self.ensure_node(tokens[TARGET])
node = parse_result_to_dsl(tokens)
self.graph.add_node_from_data(node)
return node
def handle_translocation_illegal(self, line: str, position: int,
tokens: ParseResults) -> None:
"""Handle a malformed translocation."""
raise MalformedTranslocationWarning(self.get_line_number(), line,
position, tokens)
class BelParser(BaseParser):
"""Build a parser backed by a given dictionary of namespaces"""
def __init__(self,
graph,
namespace_dict=None,
annotation_dict=None,
namespace_regex=None,
annotation_regex=None,
allow_naked_names=False,
allow_nested=False,
allow_unqualified_translocations=False,
citation_clearing=True,
no_identifier_validation=False,
autostreamline=True,
required_annotations=None):
"""
:param pybel.BELGraph graph: The BEL Graph to use to store the network
:param namespace_dict: A dictionary of {namespace: {name: encoding}}. Delegated to
:class:`pybel.parser.parse_identifier.IdentifierParser`
:type namespace_dict: Optional[dict[str,dict[str,str]]]
:param annotation_dict: A dictionary of {annotation: set of values}. Delegated to
:class:`pybel.parser.ControlParser`
:rype annotation_dict: Optional[dict[str,set[str]]]
:param namespace_regex: A dictionary of {namespace: regular expression strings}. Delegated to
:class:`pybel.parser.parse_identifier.IdentifierParser`
:type namespace_regex: Optional[dict[str,str]]
:param annotation_regex: A dictionary of {annotation: regular expression strings}. Delegated to
:class:`pybel.parser.ControlParser`
:type annotation_regex: Optional[dict[str,str]]
:param bool allow_naked_names: If true, turn off naked namespace failures. Delegated to
:class:`pybel.parser.parse_identifier.IdentifierParser`
:param bool allow_nested: If true, turn off nested statement failures. Delegated to
:class:`pybel.parser.parse_identifier.IdentifierParser`
:param bool allow_unqualified_translocations: If true, allow translocations without TO and FROM clauses.
:param bool citation_clearing: Should :code:`SET Citation` statements clear evidence and all annotations?
Delegated to :class:`pybel.parser.ControlParser`
:param bool autostreamline: Should the parser be streamlined on instantiation?
:param Optional[list[str]] required_annotations: Optional list of required annotations
"""
self.graph = graph
self.allow_nested = allow_nested
self.control_parser = ControlParser(
annotation_dict=annotation_dict,
annotation_regex=annotation_regex,
citation_clearing=citation_clearing,
required_annotations=required_annotations,
)
if no_identifier_validation:
self.identifier_parser = IdentifierParser(
allow_naked_names=allow_naked_names, )
else:
self.identifier_parser = IdentifierParser(
allow_naked_names=allow_naked_names,
namespace_dict=namespace_dict,
namespace_regex=namespace_regex,
)
identifier = Group(self.identifier_parser.language)(IDENTIFIER)
ungrouped_identifier = self.identifier_parser.language
# 2.2 Abundance Modifier Functions
#: `2.2.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_protein_modifications>`_
self.pmod = ProteinModificationParser(self.identifier_parser).language
#: `2.2.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_variant_var>`_
self.variant = VariantParser().language
#: `2.2.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_proteolytic_fragments>`_
self.fragment = FragmentParser().language
#: `2.2.4 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_cellular_location>`_
self.location = LocationParser(self.identifier_parser).language
opt_location = Optional(WCW + self.location)
#: DEPRECATED: `2.2.X Amino Acid Substitutions <http://openbel.org/language/version_1.0/bel_specification_version_1.0.html#_amino_acid_substitutions>`_
self.psub = ProteinSubstitutionParser().language
#: DEPRECATED: `2.2.X Sequence Variations <http://openbel.org/language/version_1.0/bel_specification_version_1.0.html#_sequence_variations>`_
self.gsub = GeneSubstitutionParser().language
#: DEPRECATED
#: `Truncated proteins <http://openbel.org/language/version_1.0/bel_specification_version_1.0.html#_truncated_proteins>`_
self.trunc = TruncationParser().language
#: PyBEL BEL Specification variant
self.gmod = GeneModificationParser(
).language # FIXME add identifier parser to this
# 2.6 Other Functions
#: `2.6.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_fusion_fus>`_
self.fusion = FusionParser(self.identifier_parser).language
# 2.1 Abundance Functions
#: `2.1.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XcomplexA>`_
self.general_abundance = general_abundance_tags + nest(
ungrouped_identifier + opt_location)
self.gene_modified = ungrouped_identifier + Optional(
WCW + delimitedList(Group(self.variant | self.gsub | self.gmod))
(VARIANTS))
self.gene_fusion = Group(self.fusion)(FUSION)
self.gene_fusion_legacy = Group(build_legacy_fusion(identifier,
'c'))(FUSION)
self.gene = gene_tag + nest(
MatchFirst([
self.gene_fusion, self.gene_fusion_legacy, self.gene_modified
]) + opt_location)
"""`2.1.4 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XgeneA>`_"""
self.mirna_modified = ungrouped_identifier + Optional(
WCW + delimitedList(Group(self.variant))(VARIANTS)) + opt_location
self.mirna = mirna_tag + nest(self.mirna_modified)
"""`2.1.5 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XmicroRNAA>`_"""
self.protein_modified = ungrouped_identifier + Optional(
WCW + delimitedList(
Group(
MatchFirst([
self.pmod, self.variant, self.fragment, self.psub,
self.trunc
])))(VARIANTS))
self.protein_fusion = Group(self.fusion)(FUSION)
self.protein_fusion_legacy = Group(build_legacy_fusion(
identifier, 'p'))(FUSION)
self.protein = protein_tag + nest(
MatchFirst([
self.protein_fusion,
self.protein_fusion_legacy,
self.protein_modified,
]) + opt_location)
"""`2.1.6 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XproteinA>`_"""
self.rna_modified = ungrouped_identifier + Optional(
WCW + delimitedList(Group(self.variant))(VARIANTS))
self.rna_fusion = Group(self.fusion)(FUSION)
self.rna_fusion_legacy = Group(build_legacy_fusion(identifier,
'r'))(FUSION)
self.rna = rna_tag + nest(
MatchFirst([
self.rna_fusion,
self.rna_fusion_legacy,
self.rna_modified,
]) + opt_location)
"""`2.1.7 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XrnaA>`_"""
self.single_abundance = MatchFirst([
self.general_abundance, self.gene, self.mirna, self.protein,
self.rna
])
#: `2.1.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XcomplexA>`_
self.complex_singleton = complex_tag + nest(ungrouped_identifier +
opt_location)
self.complex_list = complex_tag + nest(
delimitedList(Group(self.single_abundance
| self.complex_singleton))(MEMBERS) +
opt_location)
self.complex_abundances = self.complex_list | self.complex_singleton
# Definition of all simple abundances that can be used in a composite abundance
self.simple_abundance = self.complex_abundances | self.single_abundance
self.simple_abundance.setParseAction(self.check_function_semantics)
#: `2.1.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XcompositeA>`_
self.composite_abundance = composite_abundance_tag + nest(
delimitedList(Group(self.simple_abundance))(MEMBERS) +
opt_location)
self.abundance = self.simple_abundance | self.composite_abundance
# 2.4 Process Modifier Function
# backwards compatibility with BEL v1.0
molecular_activity_default = oneOf(list(
language.activity_labels)).setParseAction(
handle_molecular_activity_default)
#: `2.4.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XmolecularA>`_
self.molecular_activity = molecular_activity_tags + nest(
molecular_activity_default | self.identifier_parser.language)
# 2.3 Process Functions
#: `2.3.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_biologicalprocess_bp>`_
self.biological_process = biological_process_tag + nest(
ungrouped_identifier)
#: `2.3.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_pathology_path>`_
self.pathology = pathology_tag + nest(ungrouped_identifier)
self.bp_path = self.biological_process | self.pathology
self.bp_path.setParseAction(self.check_function_semantics)
self.activity_standard = activity_tag + nest(
Group(self.simple_abundance)(TARGET) +
Optional(WCW + Group(self.molecular_activity)(EFFECT)))
activity_legacy_tags = oneOf(language.activities)(MODIFIER)
self.activity_legacy = activity_legacy_tags + nest(
Group(self.simple_abundance)(TARGET))
self.activity_legacy.setParseAction(handle_activity_legacy)
self.activity = self.activity_standard | self.activity_legacy
"""`2.3.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#Xactivity>`_"""
self.process = self.bp_path | self.activity
# 2.5 Transformation Functions
from_loc = Suppress(FROM_LOC) + nest(identifier(FROM_LOC))
to_loc = Suppress(TO_LOC) + nest(identifier(TO_LOC))
self.cell_secretion = cell_secretion_tag + nest(
Group(self.simple_abundance)(TARGET))
self.cell_surface_expression = cell_surface_expression_tag + nest(
Group(self.simple_abundance)(TARGET))
self.translocation_standard = nest(
Group(self.simple_abundance)(TARGET) + WCW +
Group(from_loc + WCW + to_loc)(EFFECT))
self.translocation_legacy = nest(
Group(self.simple_abundance)(TARGET) + WCW +
Group(identifier(FROM_LOC) + WCW + identifier(TO_LOC))(EFFECT))
self.translocation_legacy.addParseAction(handle_legacy_tloc)
self.translocation_unqualified = nest(
Group(self.simple_abundance)(TARGET))
if not allow_unqualified_translocations:
self.translocation_unqualified.setParseAction(
self.handle_translocation_illegal)
self.translocation = translocation_tag + MatchFirst([
self.translocation_unqualified, self.translocation_standard,
self.translocation_legacy
])
"""`2.5.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_translocations>`_"""
#: `2.5.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_degradation_deg>`_
self.degradation = degradation_tags + nest(
Group(self.simple_abundance)(TARGET))
#: `2.5.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_reaction_rxn>`_
self.reactants = Suppress(REACTANTS) + nest(
delimitedList(Group(self.simple_abundance)))
self.products = Suppress(PRODUCTS) + nest(
delimitedList(Group(self.simple_abundance)))
self.reaction = reaction_tags + nest(
Group(self.reactants)(REACTANTS),
Group(self.products)(PRODUCTS))
self.transformation = MatchFirst([
self.cell_secretion, self.cell_surface_expression,
self.translocation, self.degradation, self.reaction
])
# 3 BEL Relationships
self.bel_term = MatchFirst(
[self.transformation, self.process, self.abundance]).streamline()
# BEL Term to BEL Term Relationships
#: `3.1.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#Xincreases>`_
increases_tag = oneOf(['->', '→', 'increases'
]).setParseAction(replaceWith(INCREASES))
#: `3.1.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XdIncreases>`_
directly_increases_tag = one_of_tags(['=>', '⇒', 'directlyIncreases'],
DIRECTLY_INCREASES)
#: `3.1.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#Xdecreases>`_
decreases_tag = one_of_tags(['-|', 'decreases'], DECREASES)
#: `3.1.4 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XdDecreases>`_
directly_decreases_tag = one_of_tags(['=|', 'directlyDecreases'],
DIRECTLY_DECREASES)
#: `3.5.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_analogous>`_
analogous_tag = one_of_tags(['analogousTo'], ANALOGOUS_TO)
#: `3.1.6 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#Xcnc>`_
causes_no_change_tag = one_of_tags(['cnc', 'causesNoChange'],
CAUSES_NO_CHANGE)
#: `3.1.7 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_regulates_reg>`_
regulates_tag = one_of_tags(['reg', 'regulates'], REGULATES)
#: `3.2.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XnegCor>`_
negative_correlation_tag = one_of_tags(['neg', 'negativeCorrelation'],
NEGATIVE_CORRELATION)
#: `3.2.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#XposCor>`_
positive_correlation_tag = one_of_tags(['pos', 'positiveCorrelation'],
POSITIVE_CORRELATION)
#: `3.2.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#Xassociation>`_
association_tag = one_of_tags(['--', 'association'], ASSOCIATION)
#: `3.3.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_orthologous>`_
orthologous_tag = one_of_tags(['orthologous'], ORTHOLOGOUS)
#: `3.4.5 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_isa>`_
is_a_tag = Keyword(IS_A)
#: PyBEL Variants
equivalent_tag = one_of_tags(['eq', EQUIVALENT_TO], EQUIVALENT_TO)
partof_tag = Keyword(PART_OF)
self.bel_to_bel_relations = [
association_tag,
increases_tag,
decreases_tag,
positive_correlation_tag,
negative_correlation_tag,
causes_no_change_tag,
orthologous_tag,
is_a_tag,
equivalent_tag,
partof_tag,
directly_increases_tag,
directly_decreases_tag,
analogous_tag,
regulates_tag,
]
self.bel_to_bel = triple(self.bel_term,
MatchFirst(self.bel_to_bel_relations),
self.bel_term)
# Mixed Relationships
#: `3.1.5 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_ratelimitingstepof>`_
rate_limit_tag = oneOf(['rateLimitingStepOf']).setParseAction(
replaceWith(RATE_LIMITING_STEP_OF))
self.rate_limit = triple(
MatchFirst(
[self.biological_process, self.activity, self.transformation]),
rate_limit_tag, self.biological_process)
#: `3.4.6 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_subprocessof>`_
subprocess_of_tag = oneOf(['subProcessOf']).setParseAction(
replaceWith(SUBPROCESS_OF))
self.subprocess_of = triple(
MatchFirst([self.process, self.activity, self.transformation]),
subprocess_of_tag, self.process)
#: `3.3.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_transcribedto>`_
transcribed_tag = oneOf([':>', 'transcribedTo'
]).setParseAction(replaceWith(TRANSCRIBED_TO))
self.transcribed = triple(self.gene, transcribed_tag, self.rna)
#: `3.3.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_translatedto>`_
translated_tag = oneOf(['>>', 'translatedTo'
]).setParseAction(replaceWith(TRANSLATED_TO))
self.translated = triple(self.rna, translated_tag, self.protein)
#: `3.4.1 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_hasmember>`_
has_member_tag = oneOf(['hasMember'
]).setParseAction(replaceWith(HAS_MEMBER))
self.has_member = triple(self.abundance, has_member_tag,
self.abundance)
#: `3.4.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_hasmembers>`_
self.abundance_list = Suppress('list') + nest(
delimitedList(Group(self.abundance)))
has_members_tag = oneOf(['hasMembers'])
self.has_members = triple(self.abundance, has_members_tag,
self.abundance_list)
self.has_members.setParseAction(self.handle_has_members)
has_components_tag = oneOf(['hasComponents'])
self.has_components = triple(self.abundance, has_components_tag,
self.abundance_list)
self.has_components.setParseAction(self.handle_has_components)
self.has_list = self.has_members | self.has_components
# `3.4.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_hascomponent>`_
has_component_tag = oneOf(['hasComponent']).setParseAction(
replaceWith(HAS_COMPONENT))
self.has_component = triple(
self.complex_abundances | self.composite_abundance,
has_component_tag, self.abundance)
#: `3.5.2 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_biomarkerfor>`_
biomarker_tag = oneOf(['biomarkerFor'
]).setParseAction(replaceWith(BIOMARKER_FOR))
#: `3.5.3 <http://openbel.org/language/version_2.0/bel_specification_version_2.0.html#_prognosticbiomarkerfor>`_
prognostic_biomarker_tag = oneOf([
'prognosticBiomarkerFor'
]).setParseAction(replaceWith(PROGONSTIC_BIOMARKER_FOR))
biomarker_tags = biomarker_tag | prognostic_biomarker_tag
self.biomarker = triple(self.bel_term, biomarker_tags, self.process)
has_variant_tags = oneOf(['hasVariant'
]).setParseAction(replaceWith(HAS_VARIANT))
self.has_variant_relation = triple(self.abundance, has_variant_tags,
self.abundance)
has_reactant_tags = oneOf(['hasReactant'
]).setParseAction(replaceWith(HAS_REACTANT))
has_product_tags = oneOf(['hasProduct'
]).setParseAction(replaceWith(HAS_PRODUCT))
part_of_reaction_tags = has_reactant_tags | has_product_tags
self.part_of_reaction = triple(self.reaction, part_of_reaction_tags,
self.abundance)
self.relation = MatchFirst([
self.bel_to_bel,
# self.has_member,
# self.has_component,
self.subprocess_of,
self.rate_limit,
self.biomarker,
self.transcribed,
self.translated,
# self.has_variant_relation,
# self.part_of_reaction,
])
self.relation.setParseAction(self._handle_relation_harness)
self.unqualified_relation = MatchFirst([
self.has_member, self.has_component, self.has_variant_relation,
self.part_of_reaction
])
self.unqualified_relation.setParseAction(
self.handle_unqualified_relation)
#: 3.1 Causal Relationships - nested. Not enabled by default.
causal_relation_tags = MatchFirst([
increases_tag, decreases_tag, directly_decreases_tag,
directly_increases_tag
])
self.nested_causal_relationship = triple(
self.bel_term, causal_relation_tags,
nest(triple(self.bel_term, causal_relation_tags, self.bel_term)))
self.nested_causal_relationship.setParseAction(
self.handle_nested_relation)
self.label_relationship = And([
Group(self.bel_term)(SUBJECT),
Suppress('labeled'),
quote(OBJECT)
])
self.label_relationship.setParseAction(self.handle_label_relation)
# has_members is handled differently from all other relations becuase it gets distrinbuted
self.relation = MatchFirst([
self.has_list,
self.nested_causal_relationship,
self.relation,
self.unqualified_relation,
self.label_relationship,
])
self.singleton_term = (self.bel_term + StringEnd()).setParseAction(
self.handle_term)
self.statement = self.relation | self.singleton_term
self.language = self.control_parser.language | self.statement
self.language.setName('BEL')
super(BelParser, self).__init__(self.language,
streamline=autostreamline)
@property
def namespace_dict(self):
"""The dictionary of {namespace: {name: encoding}} stored in the internal identifier parser
:rtype: dict[str,dict[str,str]]
"""
return self.identifier_parser.namespace_dict
@property
def namespace_regex(self):
"""The dictionary of {namespace keyword: compiled regular expression} stored the internal identifier parser
:rtype: dict[str,re]
"""
return self.identifier_parser.namespace_regex_compiled
@property
def annotation_dict(self):
"""A dictionary of annotations to their set of values
:rtype: dict[str,set[str]]
"""
return self.control_parser.annotation_dict
@property
def annotation_regex(self):
"""A dictionary of annotations defined by regular expressions {annotation keyword: string regular expression}
:rtype: dict[str,str]
"""
return self.control_parser.annotation_regex
@property
def allow_naked_names(self):
"""Should naked names be parsed, or should errors be thrown?
:rtype: bool
"""
return self.identifier_parser.allow_naked_names
def get_annotations(self):
"""Get current annotations in this parser
:rtype: dict
"""
return self.control_parser.get_annotations()
def clear(self):
"""Clears the graph and all control parser data (current citation, annotations, and statement group)"""
self.graph.clear()
self.control_parser.clear()
def handle_nested_relation(self, line, position, tokens):
"""Handles nested statements. If :code:`allow_nested` is False, raises a warning.
:param str line: The line being parsed
:param int position: The position in the line being parsed
:param pyparsing.ParseResult tokens: The tokens from PyParsing
:raises: NestedRelationWarning
"""
if not self.allow_nested:
raise NestedRelationWarning(self.line_number, line, position)
self._handle_relation_harness(
line, position, {
SUBJECT: tokens[SUBJECT],
RELATION: tokens[RELATION],
OBJECT: tokens[OBJECT][SUBJECT]
})
self._handle_relation_harness(
line, position, {
SUBJECT: tokens[OBJECT][SUBJECT],
RELATION: tokens[OBJECT][RELATION],
OBJECT: tokens[OBJECT][OBJECT]
})
return tokens
def check_function_semantics(self, line, position, tokens):
"""Raises an exception if the function used on the tokens is wrong
:param str line: The line being parsed
:param int position: The position in the line being parsed
:param pyparsing.ParseResult tokens: The tokens from PyParsing
:raises: InvalidFunctionSemantic
"""
if self.namespace_dict is None or NAMESPACE not in tokens:
return tokens
namespace, name = tokens[NAMESPACE], tokens[NAME]
if namespace in self.namespace_regex:
return tokens
if self.allow_naked_names and tokens[
NAMESPACE] == DIRTY: # Don't check dirty names in lenient mode
return tokens
valid_functions = set(
itt.chain.from_iterable(
belns_encodings[k]
for k in self.namespace_dict[namespace][name]))
if tokens[FUNCTION] not in valid_functions:
raise InvalidFunctionSemantic(self.line_number, line, position,
tokens[FUNCTION], namespace, name,
valid_functions)
return tokens
def handle_term(self, line, position, tokens):
"""Handles BEL terms (the subject and object of BEL relations)
:param str line: The line being parsed
:param int position: The position in the line being parsed
:param pyparsing.ParseResult tokens: The tokens from PyParsing
"""
self.ensure_node(tokens)
return tokens
def _handle_list_helper(self, tokens, relation):
"""Provides the functionality for :meth:`handle_has_members` and :meth:`handle_has_components`"""
parent_node_tuple, parent_node_attr = self.ensure_node(tokens[0])
for child_tokens in tokens[2]:
child_node_tuple, child_node_attr = self.ensure_node(child_tokens)
self.graph.add_unqualified_edge(parent_node_tuple,
child_node_tuple, relation)
return tokens
def handle_has_members(self, line, position, tokens):
"""Handles list relations like ``p(X) hasMembers list(p(Y), p(Z), ...)``
:param str line: The line being parsed
:param int position: The position in the line being parsed
:param pyparsing.ParseResult tokens: The tokens from PyParsing
"""
return self._handle_list_helper(tokens, HAS_MEMBER)
def handle_has_components(self, line, position, tokens):
"""Handles list relations like ``p(X) hasComponents list(p(Y), p(Z), ...)``
:param str line: The line being parsed
:param int position: The position in the line being parsed
:param pyparsing.ParseResult tokens: The tokens from PyParsing
"""
return self._handle_list_helper(tokens, HAS_COMPONENT)
def _add_qualified_edge_helper(self, u, v, relation, annotations,
subject_modifier, object_modifier):
"""Adds a qualified edge from the internal aspects of the parser"""
self.graph.add_qualified_edge(
u,
v,
relation=relation,
evidence=self.control_parser.evidence,
citation=self.control_parser.citation.copy(),
annotations=annotations,
subject_modifier=subject_modifier,
object_modifier=object_modifier,
**{LINE: self.line_number})
def _add_qualified_edge(self, u, v, relation, annotations,
subject_modifier, object_modifier):
"""Adds an edge, then adds the opposite direction edge if it should"""
self._add_qualified_edge_helper(
u,
v,
relation=relation,
annotations=annotations,
subject_modifier=subject_modifier,
object_modifier=object_modifier,
)
if relation in TWO_WAY_RELATIONS:
self._add_qualified_edge_helper(
v,
u,
relation=relation,
annotations=annotations,
object_modifier=subject_modifier,
subject_modifier=object_modifier,
)
def _handle_relation(self, tokens):
"""A policy in which all annotations are stored as sets, including single annotations
:param pyparsing.ParseResult tokens: The tokens from PyParsing
"""
subject_node_tuple, _ = self.ensure_node(tokens[SUBJECT])
object_node_tuple, _ = self.ensure_node(tokens[OBJECT])
subject_modifier = modifier_po_to_dict(tokens[SUBJECT])
object_modifier = modifier_po_to_dict(tokens[OBJECT])
annotations = {
annotation_name: ({ae: True
for ae in annotation_entry} if isinstance(
annotation_entry, set) else {
annotation_entry: True
})
for annotation_name, annotation_entry in
self.control_parser.annotations.items()
}
self._add_qualified_edge(
subject_node_tuple,
object_node_tuple,
relation=tokens[RELATION],
annotations=annotations,
subject_modifier=subject_modifier,
object_modifier=object_modifier,
)
def _handle_relation_harness(self, line, position, tokens):
"""Handles BEL relations based on the policy specified on instantiation. Note: this can't be changed after
instantiation!
:param str line: The line being parsed
:param int position: The position in the line being parsed
:param pyparsing.ParseResult tokens: The tokens from PyParsing
"""
if not self.control_parser.citation:
raise MissingCitationException(self.line_number, line, position)
if not self.control_parser.evidence:
raise MissingSupportWarning(self.line_number, line, position)
missing_required_annotations = self.control_parser.get_missing_required_annotations(
)
if missing_required_annotations:
raise MissingAnnotationWarning(self.line_number, line, position,
missing_required_annotations)
self._handle_relation(tokens)
return tokens
def handle_unqualified_relation(self, line, position, tokens):
"""Handles unqualified relations
:param str line: The line being parsed
:param int position: The position in the line being parsed
:param pyparsing.ParseResult tokens: The tokens from PyParsing
"""
subject_node_tuple, _ = self.ensure_node(tokens[SUBJECT])
object_node_tuple, _ = self.ensure_node(tokens[OBJECT])
rel = tokens[RELATION]
self.graph.add_unqualified_edge(subject_node_tuple, object_node_tuple,
rel)
def handle_label_relation(self, line, position, tokens):
"""Handles statements like ``p(X) label "Label for X"``
:param str line: The line being parsed
:param int position: The position in the line being parsed
:param pyparsing.ParseResult tokens: The tokens from PyParsing
:raises: RelabelWarning
"""
subject_node_tuple, _ = self.ensure_node(tokens[SUBJECT])
description = tokens[OBJECT]
if self.graph.has_node_description(subject_node_tuple):
raise RelabelWarning(
line_number=self.line_number,
line=line,
position=position,
node=self.graph.node,
old_label=self.graph.get_node_description(subject_node_tuple),
new_label=description)
self.graph.set_node_description(subject_node_tuple, description)
def ensure_node(self, tokens):
"""Turns parsed tokens into canonical node name and makes sure its in the graph
:param pyparsing.ParseResult tokens: Tokens from PyParsing
:return: A pair of the PyBEL node tuple and the PyBEL node data dictionary
:rtype: tuple[tuple, dict]
"""
if MODIFIER in tokens:
return self.ensure_node(tokens[TARGET])
node_attr_dict = po_to_dict(tokens)
node_tuple = self.graph.add_node_from_data(node_attr_dict)
return node_tuple, node_attr_dict
def handle_translocation_illegal(self, line, position, tokens):
raise MalformedTranslocationWarning(self.line_number, line, position,
tokens)
def __init__(self):
self.json_query = {'query': {}, 'and': [], 'or': []}
self.tokens = None
#--------------------------------------------------------------------------------------
# <integer> ::= 0-9
# <double> ::= 0-9 ('.' 0-9)
# <number> ::= <integer> | <double>
#--------------------------------------------------------------------------------------
integer = Regex(r'-?[0-9]+') # Word matches space for some reason
double = Regex(r'-?[0-9]+.?[0-9]*')
number = double | integer
#--------------------------------------------------------------------------------------
# <python-string> ::= (String surrounded by double-quotes)
# <wildcard-string> ::= <python-string>
# <limited-string> ::= '"' a..z A..Z 9..9 _ . '"' (alpha nums and ._ surrounded by double quotes)
# <field> ::= <limited-string> | "*"
# <coords> ::= "LAT" <number> "LON" <number>
# <units> ::= ('km' | 'mi' | 'nm')
# <distance> ::= REGEX(([0-9]*\.?[0-9]*)(km|mi|nm)?)
#--------------------------------------------------------------------------------------
python_string = quotedString.setParseAction(removeQuotes)
wildcard_string = python_string
limited_string = Regex(
r'("(?:[a-zA-Z0-9_\.])*"|\'(?:[a-zA-Z0-9_\.]*)\')').setParseAction(
removeQuotes)
field = limited_string ^ CaselessLiteral('"*"').setParseAction(
removeQuotes)
coords = CaselessLiteral("LAT") + number + CaselessLiteral(
"LON") + number
units = CaselessLiteral('km') | CaselessLiteral('mi')
distance = number + units
distance.setParseAction(lambda x: self.frame.update({
'dist': float(x[0]),
'units': x[1]
}))
#--------------------------------------------------------------------------------------
# Date
#--------------------------------------------------------------------------------------
date = python_string
#--------------------------------------------------------------------------------------
# <query-filter> ::= "FILTER" <python-string>
# <index-name> ::= <python-string>
# <resource-id> ::= '"' a..z A..Z 0..9 $ _ -'"' (alpha nums surrounded by double quotes)
# <collection-id> ::= <resource-id>
#--------------------------------------------------------------------------------------
query_filter = CaselessLiteral("FILTER") + python_string
# Add the filter to the frame object
query_filter.setParseAction(
lambda x: self.frame.update({'filter': x[1]}))
index_name = MatchFirst(python_string)
# Add the index to the frame object
index_name.setParseAction(lambda x: self.frame.update({'index': x[0]}))
resource_id = Regex(
r'("(?:[a-zA-Z0-9\$_-])*"|\'(?:[a-zA-Z0-9\$_-]*)\')'
).setParseAction(removeQuotes)
collection_id = resource_id
#--------------------------------------------------------------------------------------
# <from-statement> ::= "FROM" <number>
# <to-statement> ::= "TO" <number>
#--------------------------------------------------------------------------------------
from_statement = CaselessLiteral("FROM") + number
from_statement.setParseAction(
lambda x: self.frame.update({'from': x[1]}))
to_statement = CaselessLiteral("TO") + number
to_statement.setParseAction(lambda x: self.frame.update({'to': x[1]}))
#--------------------------------------------------------------------------------------
# <date-from-statement> ::= "FROM" <date>
# <date-to-statement> ::= "TO" <date>
#--------------------------------------------------------------------------------------
date_from_statement = CaselessLiteral("FROM") + date
date_from_statement.setParseAction(
lambda x: self.frame.update({'from': x[1]}))
date_to_statement = CaselessLiteral("TO") + date
date_to_statement.setParseAction(
lambda x: self.frame.update({'to': x[1]}))
#--------------------------------------------------------------------------------------
# <time-query> ::= "TIME FROM" <date> "TO" <date>
#--------------------------------------------------------------------------------------
time_query = CaselessLiteral("TIME") + Optional(
date_from_statement) + Optional(date_to_statement)
time_query.setParseAction(lambda x: self.time_frame())
# time.mktime(dateutil.parser.parse(x[2])), 'to':time.mktime(dateutil.parser.parse(x[4]))}}))
#--------------------------------------------------------------------------------------
# <time-bounds> ::= "TIMEBOUNDS" <from-statement> <to-statement>
#--------------------------------------------------------------------------------------
time_bounds = CaselessLiteral(
"TIMEBOUNDS") + date_from_statement + date_to_statement
time_bounds.setParseAction(lambda x: self.time_bounds_frame())
#--------------------------------------------------------------------------------------
# <vertical-bounds> ::= "VERTICAL" <from-statement> <to-statement>
#--------------------------------------------------------------------------------------
vertical_bounds = CaselessLiteral(
"VERTICAL") + from_statement + to_statement
vertical_bounds.setParseAction(lambda x: self.vertical_bounds_frame())
#--------------------------------------------------------------------------------------
# <range-query> ::= "VALUES" [<from-statement>] [<to-statement>]
#--------------------------------------------------------------------------------------
range_query = CaselessLiteral("VALUES") + Optional(
from_statement) + Optional(to_statement)
# Add the range to the frame object
range_query.setParseAction(lambda x: self.range_frame())
#--------------------------------------------------------------------------------------
# <geo-distance> ::= "DISTANCE" <distance> "FROM" <coords>
# <geo-bbox> ::= "BOX" "TOP-LEFT" <coords> "BOTTOM-RIGHT" <coords>
#--------------------------------------------------------------------------------------
geo_distance = CaselessLiteral(
"DISTANCE") + distance + CaselessLiteral("FROM") + coords
geo_distance.setParseAction(
lambda x: self.frame.update({
'lat': float(x[5]),
'lon': float(x[7])
}))
geo_bbox = CaselessLiteral("BOX") + CaselessLiteral(
"TOP-LEFT") + coords + CaselessLiteral("BOTTOM-RIGHT") + coords
geo_bbox.setParseAction(lambda x: self.frame.update({
'top_left': [float(x[5]), float(x[3])],
'bottom_right': [float(x[10]), float(x[8])]
}))
#--------------------------------------------------------------------------------------
# <field-query> ::= <wildcard-string>
# <term-query> ::= "IS" <field-query>
# <fuzzy-query> ::= "LIKE" <field-query>
# <match-query> ::= "MATCH" <field-query>
# <geo-query> ::= "GEO" ( <geo-distance> | <geo-bbox> )
#--------------------------------------------------------------------------------------
field_query = wildcard_string
term_query = CaselessLiteral("IS") + field_query
term_query.setParseAction(lambda x: self.frame.update({'value': x[1]}))
geo_query = CaselessLiteral("GEO") + (geo_distance | geo_bbox)
fuzzy_query = CaselessLiteral("LIKE") + field_query
fuzzy_query.setParseAction(
lambda x: self.frame.update({'fuzzy': x[1]}))
match_query = CaselessLiteral("MATCH") + field_query
match_query.setParseAction(
lambda x: self.frame.update({'match': x[1]}))
#--------------------------------------------------------------------------------------
# <limit-parameter> ::= "LIMIT" <integer>
# <depth-parameter> ::= "DEPTH" <integer>
# <order-parameter> ::= "ORDER" "BY" <limited-string>
# <offset-parameter> ::= "SKIP" <integer>
# <query-parameter> ::= <order-paramater> | <limit-parameter>
#--------------------------------------------------------------------------------------
limit_parameter = CaselessLiteral("LIMIT") + integer
limit_parameter.setParseAction(
lambda x: self.json_query.update({'limit': int(x[1])}))
depth_parameter = CaselessLiteral("DEPTH") + integer
depth_parameter.setParseAction(
lambda x: self.frame.update({'depth': int(x[1])}))
order_parameter = CaselessLiteral("ORDER") + CaselessLiteral(
"BY") + limited_string
order_parameter.setParseAction(
lambda x: self.json_query.update({'order': {
x[2]: 'asc'
}}))
offset_parameter = CaselessLiteral("SKIP") + integer
offset_parameter.setParseAction(
lambda x: self.json_query.update({'skip': int(x[1])}))
query_parameter = limit_parameter | order_parameter | offset_parameter
#--------------------------------------------------------------------------------------
# <search-query> ::= "SEARCH" <field> (<range-query> | <term-query> | <fuzzy-query> | <match-query> | <time-query> | <time-bounds> | <vertical-bounds> | <geo-query>) "FROM" <index-name> [<query-parameter>]*
# <collection-query> ::= "IN <collection-id>"
# <association-query> ::= "BELONGS TO" <resource-id> [ <depth-parameter> ]
# <owner-query> ::= "HAS" <resource-id> [ <depth-parameter> ]
# <query> ::= <search-query> | <association-query> | <collection-query> | <owner-query>
#--------------------------------------------------------------------------------------
search_query = CaselessLiteral("SEARCH") + field + (
range_query | term_query | fuzzy_query | match_query
| vertical_bounds | time_bounds | time_query
| geo_query) + CaselessLiteral("FROM") + index_name
# Add the field to the frame object
search_query.setParseAction(
lambda x: self.frame.update({'field': x[1]}))
collection_query = CaselessLiteral("IN") + collection_id
collection_query.setParseAction(
lambda x: self.frame.update({'collection': x[1]}))
association_query = CaselessLiteral("BELONGS") + CaselessLiteral(
"TO") + resource_id + Optional(depth_parameter)
# Add the association to the frame object
association_query.setParseAction(
lambda x: self.frame.update({'association': x[2]}))
owner_query = CaselessLiteral("HAS") + resource_id + Optional(
depth_parameter)
owner_query.setParseAction(
lambda x: self.frame.update({'owner': x[1]}))
query = search_query | association_query | collection_query | owner_query
#--------------------------------------------------------------------------------------
# <primary-query> ::= <query> [<query-filter>]
# <atom> ::= <query>
# <intersection> ::= "AND" <atom>
# <union> ::= "OR" <atom>
# <sentence> ::= <primary-query> [<intersection>]* [<union>]*
#--------------------------------------------------------------------------------------
primary_query = query + Optional(query_filter)
# Set the primary query on the json_query to the frame and clear the frame
primary_query.setParseAction(lambda x: self.push_frame())
atom = query
intersection = CaselessLiteral("AND") + atom
# Add an AND operation to the json_query and clear the frame
intersection.setParseAction(lambda x: self.and_frame())
union = CaselessLiteral("OR") + atom
# Add an OR operation to the json_query and clear the frame
union.setParseAction(lambda x: self.or_frame())
self.sentence = primary_query + (intersection ^ union) * (
0, None) + query_parameter * (0, None)
def evaluator(variables, functions, string, cs=False):
"""
Evaluate an expression. Variables are passed as a dictionary
from string to value. Unary functions are passed as a dictionary
from string to function. Variables must be floats.
cs: Case sensitive
"""
all_variables = copy.copy(DEFAULT_VARIABLES)
all_functions = copy.copy(DEFAULT_FUNCTIONS)
all_variables.update(variables)
all_functions.update(functions)
if not cs:
string_cs = string.lower()
all_functions = lower_dict(all_functions)
all_variables = lower_dict(all_variables)
CasedLiteral = CaselessLiteral
else:
string_cs = string
CasedLiteral = Literal
check_variables(string_cs,
set(all_variables.keys() + all_functions.keys()))
if string.strip() == "":
return float('nan')
# SI suffixes and percent
number_suffix = MatchFirst([Literal(k) for k in SUFFIXES.keys()])
plus_minus = Literal('+') | Literal('-')
times_div = Literal('*') | Literal('/')
number_part = Word(nums)
# 0.33 or 7 or .34 or 16.
inner_number = (number_part + Optional("." + Optional(number_part))) | (
"." + number_part)
# by default pyparsing allows spaces between tokens--Combine prevents that
inner_number = Combine(inner_number)
# 0.33k or -17
number = (
inner_number +
Optional(CaselessLiteral("E") + Optional(plus_minus) + number_part) +
Optional(number_suffix))
number.setParseAction(number_parse_action) # Convert to number
# Predefine recursive variables
expr = Forward()
# Handle variables passed in.
# E.g. if we have {'R':0.5}, we make the substitution.
# We sort the list so that var names (like "e2") match before
# mathematical constants (like "e"). This is kind of a hack.
all_variables_keys = sorted(all_variables.keys(), key=len, reverse=True)
varnames = MatchFirst([CasedLiteral(k) for k in all_variables_keys])
varnames.setParseAction(lambda x: [all_variables[k] for k in x])
# if all_variables were empty, then pyparsing wants
# varnames = NoMatch()
# this is not the case, as all_variables contains the defaults
# Same thing for functions.
all_functions_keys = sorted(all_functions.keys(), key=len, reverse=True)
funcnames = MatchFirst([CasedLiteral(k) for k in all_functions_keys])
function = funcnames + Suppress("(") + expr + Suppress(")")
function.setParseAction(lambda x: [all_functions[x[0]](x[1])])
atom = number | function | varnames | Suppress("(") + expr + Suppress(")")
# Do the following in the correct order to preserve order of operation
pow_term = atom + ZeroOrMore(Suppress("^") + atom)
pow_term.setParseAction(exp_parse_action) # 7^6
par_term = pow_term + ZeroOrMore(Suppress('||') + pow_term) # 5k || 4k
par_term.setParseAction(parallel)
prod_term = par_term + ZeroOrMore(times_div + par_term) # 7 * 5 / 4 - 3
prod_term.setParseAction(prod_parse_action)
sum_term = Optional(plus_minus) + prod_term + ZeroOrMore(
plus_minus + prod_term) # -5 + 4 - 3
sum_term.setParseAction(sum_parse_action)
expr << sum_term # finish the recursion
return (expr + stringEnd).parseString(string)[0]
FLOAT_CONSTANT = Regex(r'-?\d*\.\d+')
CONSTANT = STRING_CONSTANT | FLOAT_CONSTANT | INT_CONSTANT
# All constants should be interpreted in Python format
CONSTANT.setParseAction(lambda toks: ast.literal_eval(toks[0]))
COMMA = Literal(',')
LIST = Literal('[') + CONSTANT + ZeroOrMore(COMMA + CONSTANT) + Optional(
COMMA) + Literal(']')
IDENTIFIER = MatchFirst(
[Word(alphas, alphanums),
QuotedString('[', endQuoteChar=']')])
IDENTIFIER.setParseAction(lambda toks: Where(toks[0]))
VALUE = CONSTANT | IDENTIFIER
OPERATOR = MatchFirst([
Regex(r'==?'),
Literal('!='),
Literal('<>'),
Literal('>='),
Literal('<='),
Literal('>'),
Literal('<'),
CaselessKeyword('like'),
])
def evaluator(variables, functions, string, cs=False):
"""
Evaluate an expression. Variables are passed as a dictionary
from string to value. Unary functions are passed as a dictionary
from string to function. Variables must be floats.
cs: Case sensitive
"""
all_variables = copy.copy(DEFAULT_VARIABLES)
all_functions = copy.copy(DEFAULT_FUNCTIONS)
all_variables.update(variables)
all_functions.update(functions)
if not cs:
string_cs = string.lower()
all_functions = lower_dict(all_functions)
all_variables = lower_dict(all_variables)
CasedLiteral = CaselessLiteral
else:
string_cs = string
CasedLiteral = Literal
check_variables(string_cs, set(all_variables.keys() + all_functions.keys()))
if string.strip() == "":
return float('nan')
# SI suffixes and percent
number_suffix = MatchFirst([Literal(k) for k in SUFFIXES.keys()])
plus_minus = Literal('+') | Literal('-')
times_div = Literal('*') | Literal('/')
number_part = Word(nums)
# 0.33 or 7 or .34 or 16.
inner_number = (number_part + Optional("." + Optional(number_part))) | ("." + number_part)
# by default pyparsing allows spaces between tokens--Combine prevents that
inner_number = Combine(inner_number)
# 0.33k or -17
number = (inner_number
+ Optional(CaselessLiteral("E") + Optional(plus_minus) + number_part)
+ Optional(number_suffix))
number.setParseAction(number_parse_action) # Convert to number
# Predefine recursive variables
expr = Forward()
# Handle variables passed in.
# E.g. if we have {'R':0.5}, we make the substitution.
# We sort the list so that var names (like "e2") match before
# mathematical constants (like "e"). This is kind of a hack.
all_variables_keys = sorted(all_variables.keys(), key=len, reverse=True)
varnames = MatchFirst([CasedLiteral(k) for k in all_variables_keys])
varnames.setParseAction(
lambda x: [all_variables[k] for k in x]
)
# if all_variables were empty, then pyparsing wants
# varnames = NoMatch()
# this is not the case, as all_variables contains the defaults
# Same thing for functions.
all_functions_keys = sorted(all_functions.keys(), key=len, reverse=True)
funcnames = MatchFirst([CasedLiteral(k) for k in all_functions_keys])
function = funcnames + Suppress("(") + expr + Suppress(")")
function.setParseAction(
lambda x: [all_functions[x[0]](x[1])]
)
atom = number | function | varnames | Suppress("(") + expr + Suppress(")")
# Do the following in the correct order to preserve order of operation
pow_term = atom + ZeroOrMore(Suppress("^") + atom)
pow_term.setParseAction(exp_parse_action) # 7^6
par_term = pow_term + ZeroOrMore(Suppress('||') + pow_term) # 5k || 4k
par_term.setParseAction(parallel)
prod_term = par_term + ZeroOrMore(times_div + par_term) # 7 * 5 / 4 - 3
prod_term.setParseAction(prod_parse_action)
sum_term = Optional(plus_minus) + prod_term + ZeroOrMore(plus_minus + prod_term) # -5 + 4 - 3
sum_term.setParseAction(sum_parse_action)
expr << sum_term # finish the recursion
return (expr + stringEnd).parseString(string)[0]
def _build_field_expr(field_names):
field = MatchFirst(
[CaselessKeyword(field_name) for field_name in field_names])
field.setParseAction(lambda x: Field(x[0]))
return field
# // BATTLE VERBS
RUN_AWAY = oneOf(['run', 'run away', 'flee'], True, False)
ENGAGE = oneOf(['attack', 'pounce', 'fight', 'engage'], True, False)
BLOCK = oneOf(['dodge', 'parry', 'block'], True, False)
def set_target(t):
r = [{'target': [x for x in t]}]
print(r)
return r
#// NORMALIZE OUR DIRECTIONS
NORTH.setParseAction(lambda: [
{
'direction': 'NORTH',
'key': 0
},
])
EAST.setParseAction(lambda: [
{
'direction': 'EAST',
'key': 1
},
])
SOUTH.setParseAction(lambda: [
{
'direction': 'SOUTH',
'key': 2
},
])
WEST.setParseAction(lambda: [
def __init__(self):
self.json_query = {"query": {}, "and": [], "or": []}
self.tokens = None
# --------------------------------------------------------------------------------------
# <integer> ::= 0-9
# <double> ::= 0-9 ('.' 0-9)
# <number> ::= <integer> | <double>
# --------------------------------------------------------------------------------------
integer = Regex(r"-?[0-9]+") # Word matches space for some reason
double = Regex(r"-?[0-9]+.?[0-9]*")
number = double | integer
# --------------------------------------------------------------------------------------
# <python-string> ::= (String surrounded by double-quotes)
# <wildcard-string> ::= <python-string>
# <limited-string> ::= '"' a..z A..Z 9..9 _ . '"' (alpha nums and ._ surrounded by double quotes)
# <field> ::= <limited-string> | "*"
# <coords> ::= "LAT" <number> "LON" <number>
# <units> ::= ('km' | 'mi' | 'nm')
# <distance> ::= REGEX(([0-9]*\.?[0-9]*)(km|mi|nm)?)
# --------------------------------------------------------------------------------------
python_string = quotedString.setParseAction(removeQuotes)
wildcard_string = python_string
limited_string = Regex(r'("(?:[a-zA-Z0-9_\.])*"|\'(?:[a-zA-Z0-9_\.]*)\')').setParseAction(removeQuotes)
field = limited_string ^ CaselessLiteral('"*"').setParseAction(removeQuotes)
coords = CaselessLiteral("LAT") + number + CaselessLiteral("LON") + number
units = CaselessLiteral("km") | CaselessLiteral("nm") | CaselessLiteral("mi")
distance = number + units
distance.setParseAction(lambda x: self.frame.update({"dist": float(x[0]), "units": x[1]}))
# --------------------------------------------------------------------------------------
# <query-filter> ::= "FILTER" <python-string>
# <index-name> ::= <python-string>
# <resource-id> ::= '"' a..z A..Z 0..9 '"' (alpha nums surrounded by double quotes)
# <collection-id> ::= <resource-id>
# --------------------------------------------------------------------------------------
query_filter = CaselessLiteral("FILTER") + python_string
# Add the filter to the frame object
query_filter.setParseAction(lambda x: self.frame.update({"filter": x[1]}))
index_name = MatchFirst(python_string)
# Add the index to the frame object
index_name.setParseAction(lambda x: self.frame.update({"index": x[0]}))
resource_id = Regex(r'("(?:[a-zA-Z0-9])*"|\'(?:[a-zA-Z0-9]*)\')').setParseAction(removeQuotes)
collection_id = resource_id
# --------------------------------------------------------------------------------------
# <range-query> ::= "VALUES FROM" <number> "TO" <number>
# --------------------------------------------------------------------------------------
range_query = CaselessLiteral("VALUES") + CaselessLiteral("FROM") + number + CaselessLiteral("TO") + number
# Add the range to the frame object
range_query.setParseAction(lambda x: self.frame.update({"range": {"from": float(x[2]), "to": float(x[4])}}))
# --------------------------------------------------------------------------------------
# <geo-distance> ::= "DISTANCE" <distance> "FROM" <coords>
# <geo-bbox> ::= "BOX" "TOP-LEFT" <coords> "BOTTOM-RIGHT" <coords>
# --------------------------------------------------------------------------------------
geo_distance = CaselessLiteral("DISTANCE") + distance + CaselessLiteral("FROM") + coords
geo_distance.setParseAction(lambda x: self.frame.update({"lat": float(x[5]), "lon": float(x[7])}))
geo_bbox = (
CaselessLiteral("BOX") + CaselessLiteral("TOP-LEFT") + coords + CaselessLiteral("BOTTOM-RIGHT") + coords
)
geo_bbox.setParseAction(
lambda x: self.frame.update(
{"top_left": [float(x[5]), float(x[3])], "bottom_right": [float(x[10]), float(x[8])]}
)
)
# --------------------------------------------------------------------------------------
# <field-query> ::= <wildcard-string>
# <term-query> ::= "IS" <field-query>
# <geo-query> ::= "GEO" ( <geo-distance> | <geo-bbox> )
# --------------------------------------------------------------------------------------
field_query = wildcard_string
term_query = CaselessLiteral("IS") + field_query
# Add the term to the frame object
term_query.setParseAction(lambda x: self.frame.update({"value": x[1]}))
geo_query = CaselessLiteral("GEO") + (geo_distance | geo_bbox)
# --------------------------------------------------------------------------------------
# <limit-parameter> ::= "LIMIT" <integer>
# <depth-parameter> ::= "DEPTH" <integer>
# <order-parameter> ::= "ORDER" "BY" <limited-string>
# <offset-parameter> ::= "SKIP" <integer>
# <query-parameter> ::= <order-paramater> | <limit-parameter>
# --------------------------------------------------------------------------------------
limit_parameter = CaselessLiteral("LIMIT") + integer
limit_parameter.setParseAction(lambda x: self.frame.update({"limit": int(x[1])}))
depth_parameter = CaselessLiteral("DEPTH") + integer
depth_parameter.setParseAction(lambda x: self.frame.update({"depth": int(x[1])}))
order_parameter = CaselessLiteral("ORDER") + CaselessLiteral("BY") + limited_string
order_parameter.setParseAction(lambda x: self.frame.update({"order": {x[2]: "asc"}}))
offset_parameter = CaselessLiteral("SKIP") + integer
offset_parameter.setParseAction(lambda x: self.frame.update({"offset": int(x[1])}))
query_parameter = limit_parameter | order_parameter | offset_parameter
# --------------------------------------------------------------------------------------
# <search-query> ::= "SEARCH" <field> (<range-query> | <term-query> | <geo-query>) "FROM" <index-name> [<query-parameter>]*
# <collection-query> ::= "IN <collection-id>"
# <association-query> ::= "BELONGS TO" <resource-id> [ <depth-parameter> ]
# <query> ::= <search-query> | <association-query> | <collection-query>
# --------------------------------------------------------------------------------------
search_query = (
CaselessLiteral("SEARCH")
+ field
+ (range_query | term_query | geo_query)
+ CaselessLiteral("FROM")
+ index_name
+ query_parameter * (0, None)
)
# Add the field to the frame object
search_query.setParseAction(lambda x: self.frame.update({"field": x[1]}))
collection_query = CaselessLiteral("IN") + collection_id
collection_query.setParseAction(lambda x: self.frame.update({"collection": x[1]}))
association_query = CaselessLiteral("BELONGS") + CaselessLiteral("TO") + resource_id + Optional(depth_parameter)
# Add the association to the frame object
association_query.setParseAction(lambda x: self.frame.update({"association": x[2]}))
query = search_query | association_query | collection_query
# --------------------------------------------------------------------------------------
# <primary-query> ::= <query> [<query-filter>]
# <atom> ::= <query>
# <intersection> ::= "AND" <atom>
# <union> ::= "OR" <atom>
# <sentence> ::= <primary-query> [<intersection>]* [<union>]*
# --------------------------------------------------------------------------------------
primary_query = query + Optional(query_filter)
# Set the primary query on the json_query to the frame and clear the frame
primary_query.setParseAction(lambda x: self.push_frame())
atom = query
intersection = CaselessLiteral("AND") + atom
# Add an AND operation to the json_query and clear the frame
intersection.setParseAction(lambda x: self.and_frame())
union = CaselessLiteral("OR") + atom
# Add an OR operation to the json_query and clear the frame
union.setParseAction(lambda x: self.or_frame())
self.sentence = primary_query + (intersection ^ union) * (0, None)