def test_validate_dictionary_is_all_text(self):
validate_dictionary_is_all_text({"1": "a", "2": "b"})
with self.assertRaises(ValueError):
validate_dictionary_is_all_text({"1": "a", "2": 1})
with self.assertRaises(ValueError):
validate_dictionary_is_all_text({"1": "a", "2": 3.14})
with self.assertRaises(ValueError):
validate_dictionary_is_all_text({"1": "a", "2": object()})
soup = BeautifulSoup("<span>foo</span>", "lxml")
navstring = soup.span.string
self.assertEqual({"a": "foo"},
validate_dictionary_is_all_text({"a": navstring}))
self.assertEqual({"a": ["foo"]},
validate_dictionary_is_all_text({"a": [navstring]}))
self.assertEqual(
{"a": {
"b": "foo"
}},
validate_dictionary_is_all_text({"a": {
"b": "foo"
}}),
)
def import_by_30_unported_license_html(self, *, content, legalcode):
"""
Returns a dictionary mapping our internal keys to strings.
"""
messages = {}
raw_html = content
# Some trivial making consistent - some translators changed 'strong' to 'b'
# for some unknown reason.
raw_html = raw_html.replace("<b>", "<strong>").replace("</b>", "</strong>")
raw_html = raw_html.replace("<B>", "<strong>").replace("</B>", "</strong>")
# Parse the raw HTML to a BeautifulSoup object.
soup = BeautifulSoup(raw_html, "lxml")
messages["license_medium"] = inner_html(soup.find(id="deed-license").h2)
legalcode.title = messages["license_medium"]
legalcode.save()
deed_main_content = soup.find(id="deed-main-content")
messages["not_a_law_firm"] = nested_text(deed_main_content.blockquote)
# <h3><em>License</em></h3>
messages["license"] = nested_text(deed_main_content.h3)
# Top level paragraphs
def paragraphs_generator():
for p in direct_children_with_tag(deed_main_content, "p"):
yield p
paragraphs = paragraphs_generator()
def ols_generator():
for ol in direct_children_with_tag(deed_main_content, "ol"):
yield ol
ordered_lists = ols_generator()
# Two paragraphs of introduction
messages["par1"] = nested_text(next(paragraphs))
messages["par2"] = nested_text(next(paragraphs))
# <p><strong>1. Definitions</strong></p>
messages["definitions"] = nested_text(next(paragraphs))
# An ordered list of definitions
ol = next(ordered_lists)
for i, li in enumerate(direct_children_with_tag(ol, "li")):
nt = name_and_text(li)
name = nt["name"]
text = nt["text"]
messages[f"def{i}name"] = name
messages[f"def{i}text"] = text
# <p><strong>2. Fair Dealing Rights.</strong> Nothing ... </p>
nt = name_and_text(next(paragraphs))
messages["fair_dealing_rights"] = nt["name"]
messages["fair_dealing_rights_text"] = nt["text"]
# <p><strong>3. License Grant.</strong> Subject ... </p>
nt = name_and_text(next(paragraphs))
messages["grant"] = nt["name"]
messages["grant_text"] = nt["text"]
# another ol
ol = next(ordered_lists)
for i, li in enumerate(direct_children_with_tag(ol, "li")):
messages[f"grant{i}"] = nested_text(li)
messages["par5"] = nested_text(next(paragraphs))
# <p><strong>4. Restrictions.</strong> The ... </p>
nt = name_and_text(next(paragraphs))
messages["restrictions"] = nt["name"]
messages["restrictions_text"] = nt["text"]
ol = next(ordered_lists)
for i, li in enumerate(direct_children_with_tag(ol, "li")):
# Most of these li's just have text.
# one has a <p></p> followed by another ordered list
if li.p:
messages["restrictions avoid doubt"] = nested_text(li.p)
ol2 = li.ol
for j, li2 in enumerate(direct_children_with_tag(ol2, "li")):
nt = name_and_text(li2)
messages[f"restrictions name {i};{j}"] = nt["name"]
messages[f"restrictions text {i};{j}"] = nt["text"]
else:
messages[f"restrictions{i}"] = nested_text(li)
# <p><strong>5. Representations, Warranties and Disclaimer</strong></p>
messages["reps_and_disclaimer"] = nested_text(next(paragraphs))
messages["unless_mutual"] = nested_text(next(paragraphs))
# <p><strong>6. Limitation on Liability.</strong> EXCEPT ...</p>
nt = name_and_text(next(paragraphs))
messages["Limitation"] = nt["name"]
messages["Limitation_text"] = nt["text"]
# <p><strong>7. Termination</strong></p>
messages["termination"] = nested_text(next(paragraphs))
ol = next(ordered_lists)
for i, li in enumerate(direct_children_with_tag(ol, "li")):
messages[f"termination{i}"] = nested_text(li)
# <p><strong>8. Miscellaneous</strong></p>
messages["misc"] = nested_text(next(paragraphs))
ol = next(ordered_lists)
for i, li in enumerate(direct_children_with_tag(ol, "li")):
messages[f"misc{i}"] = nested_text(li)
# That's it for the license. The rest is disclaimer that we're handling elsewhere.
validate_dictionary_is_all_text(messages)
return messages
def import_by_40_license_html(self, *, content, legalcode):
"""
Returns a dictionary mapping our internal keys to strings.
"""
license = legalcode.license
license_code = license.license_code
language_code = legalcode.language_code
assert license.version == "4.0", f"{license.version} is not '4.0'"
assert license.license_code.startswith("by")
messages = {}
raw_html = content
# Some trivial making consistent - some translators changed 'strong' to 'b'
# for some unknown reason.
raw_html = raw_html.replace("<b>", "<strong>").replace("</b>", "</strong>")
raw_html = raw_html.replace("<B>", "<strong>").replace("</B>", "</strong>")
# Parse the raw HTML to a BeautifulSoup object.
soup = BeautifulSoup(raw_html, "lxml")
# Get the license titles and intro text.
deed_main_content = soup.find(id="deed-main-content")
messages["license_medium"] = inner_html(soup.find(id="deed-license").h2)
legalcode.title = messages["license_medium"]
legalcode.save()
messages["license_long"] = inner_html(deed_main_content.h3)
messages["license_intro"] = inner_html(
deed_main_content.h3.find_next_sibling("p")
)
# Section 1 – Definitions.
# We're going to work out a list of what definitions we expect in this license,
# and in what order.
# Start with the definitions common to all the BY 4.0 licenses
expected_definitions = [
"adapted_material",
"copyright_and_similar_rights",
"effective_technological_measures",
"exceptions_and_limitations",
"licensed_material",
"licensed_rights",
"licensor",
"share",
"sui_generis_database_rights",
"you",
]
# now insert the optional ones
def insert_after(after_this, what_to_insert):
i = expected_definitions.index(after_this)
expected_definitions.insert(i + 1, what_to_insert)
if license_code == "by-sa":
insert_after("adapted_material", "adapters_license")
insert_after("adapters_license", "by_sa_compatible_license")
insert_after("exceptions_and_limitations", "license_elements_sa")
# See https://github.com/creativecommons/creativecommons.org/issues/1153
# BY-SA 4.0 for "pt" has an extra definition. Work around for now.
if language_code == "pt":
insert_after("you", "you2")
elif license_code == "by":
insert_after("adapted_material", "adapters_license")
elif license_code == "by-nc":
insert_after("adapted_material", "adapters_license")
insert_after("licensor", "noncommercial")
elif license_code == "by-nd":
pass
elif license_code == "by-nc-nd":
insert_after("licensor", "noncommercial")
elif license_code == "by-nc-sa":
insert_after("adapted_material", "adapters_license")
insert_after("exceptions_and_limitations", "license_elements_nc_sa")
insert_after("adapters_license", "by_nc_sa_compatible_license")
insert_after("licensor", "noncommercial")
# definitions are in an "ol" that is the next sibling of the id=s1 element.
messages["s1_definitions_title"] = inner_html(soup.find(id="s1").strong)
for i, definition in enumerate(
soup.find(id="s1").find_next_siblings("ol")[0].find_all("li")
):
thing = name_and_text(definition)
defn_key = expected_definitions[i]
messages[
f"s1_definitions_{defn_key}"
] = f"""<span style="text-decoration: underline;">{thing['name']}</span> {thing['text']}"""
# Section 2 – Scope.
messages["s2_scope"] = inner_html(soup.find(id="s2").strong)
# Section 2a - License Grant
# translation of "License grant"
s2a = soup.find(id="s2a")
if s2a.strong:
messages["s2a_license_grant_title"] = inner_html(s2a.strong)
elif s2a.b:
messages["s2a_license_grant_title"] = inner_html(s2a.b)
else:
print(f"How do I handle {s2a}?")
sys.exit(1)
# s2a1: rights
messages["s2a_license_grant_intro"] = str(list(soup.find(id="s2a1"))[0]).strip()
messages["s2a_license_grant_share"] = str(
list(soup.find(id="s2a1A"))[0]
).strip()
messages["s2a_license_grant_adapted"] = str(
list(soup.find(id="s2a1B"))[0]
).strip()
# s2a2: exceptions and limitations
nt = name_and_text(soup.find(id="s2a2"))
messages[
"s2a2_license_grant_exceptions"
] = f"<strong>{nt['name']}</strong>{nt['text']}"
# s2a3: term
nt = name_and_text(soup.find(id="s2a3"))
messages[
"s2a3_license_grant_term"
] = f"<strong>{nt['name']}</strong>{nt['text']}"
# s2a4: media
nt = name_and_text(soup.find(id="s2a4"))
messages[
"s2a4_license_grant_media"
] = f"<strong>{nt['name']}</strong>{nt['text']}"
# s2a5: scope/grant/downstream
# The title is just the prefix to the list of items, which are in their
# own div, so this is slightly messy. Using the name from name_and_text
# will get us the text we want without wrappings.
nt = name_and_text(soup.find(id="s2a5"))
messages["s2a5_license_grant_downstream_title"] = nt["name"]
expected_downstreams = [
"offer",
"no_restrictions",
]
if license_code in ["by-sa", "by-nc-sa"]:
expected_downstreams.insert(1, "adapted_material")
# Process top-level "li" elements under the ol
for i, li in enumerate(
soup.find(id="s2a5").div.ol.find_all("li", recursive=False)
):
key = expected_downstreams[i]
thing = name_and_text(li)
messages[f"s2a5_license_grant_downstream_{key}_name"] = thing["name"]
messages[f"s2a5_license_grant_downstream_{key}_text"] = thing["text"]
nt = name_and_text(soup.find(id="s2a6"))
messages["s2a6_license_grant_no_endorsement_name"] = nt["name"]
messages["s2a6_license_grant_no_endorsement_text"] = nt["text"]
# s2b: other rights
s2b = soup.find(id="s2b")
if s2b.p and s2b.p.strong:
messages["s2b_other_rights_title"] = nested_text(s2b.p.strong)
elif s2b.p:
messages["s2b_other_rights_title"] = nested_text(s2b.p)
elif s2b.strong:
messages["s2b_other_rights_title"] = nested_text(s2b.strong)
else:
print(str(s2b))
raise ValueError("Where is s2b's title?")
list_items = soup.find(id="s2b").ol.find_all("li", recursive=False)
assert list_items[0].name == "li"
messages["s2b1_other_rights_moral"] = nested_text(list_items[0])
messages["s2b2_other_rights_patent"] = nested_text(list_items[1])
messages["s2b3_other_rights_waive"] = nested_text(list_items[2])
# Section 3: conditions
s3 = soup.find(id="s3")
messages["s3_conditions_title"] = nested_text(s3)
messages["s3_conditions_intro"] = nested_text(
soup.find(id="s3").find_next_sibling("p")
)
# <p id="s3"><strong>Section 3 – License Conditions.</strong></p>
#
# <p>Your exercise of the Licensed Rights is expressly made subject to the following conditions.</p>
#
# <ol type="a">
# <li id="s3a"><p><strong>Attribution</strong>.</p>
# <ol>
s3a = soup.find(id="s3a")
inside = str(inner_html(s3a))
if inside.startswith(" "): # ar translation takes liberties with whitespace
s3a = BeautifulSoup(inside.strip(), "lxml")
if s3a.p and s3a.p.strong:
messages["s3_conditions_attribution"] = nested_text(s3a.p.strong)
elif s3a.strong:
messages["s3_conditions_attribution"] = nested_text(s3a.strong)
else:
print(str(s3a))
raise ValueError("Fix s3a's attribution string")
messages["s3_conditions_if_you_share"] = text_up_to(soup.find(id="s3a1"), "ol")
messages["s3_conditions_retain_the_following"] = text_up_to(
soup.find(id="s3a1A"), "ol"
)
messages["s3a1Ai_conditions_identification"] = inner_html(
soup.find(id="s3a1Ai")
)
messages["s3a1Aii_conditions_copyright"] = inner_html(soup.find(id="s3a1Aii"))
messages["s3a1Aiii_conditions_license"] = inner_html(soup.find(id="s3a1Aiii"))
messages["s3a1Aiv_conditions_disclaimer"] = inner_html(soup.find(id="s3a1Aiv"))
messages["s3a1Av_conditions_link"] = inner_html(soup.find(id="s3a1Av"))
messages["s3a1B_conditions_modified"] = inner_html(soup.find(id="s3a1B"))
messages["s3a1C_conditions_licensed"] = inner_html(soup.find(id="s3a1C"))
messages["s3a2_conditions_satisfy"] = inner_html(soup.find(id="s3a2"))
messages["s3a3_conditions_remove"] = inner_html(soup.find(id="s3a3"))
if soup.find(id="s3a4"):
# Only present if neither SA or ND.
# OR in the NL translation of by-nc-nd, go figure...
messages["s3a4_if_you_share_adapted_material"] = nested_text(
soup.find(id="s3a4")
)
# share-alike is only in some licenses
if license_code.endswith("-sa"):
messages["sharealike_name"] = nested_text(soup.find(id="s3b").strong)
messages["sharealike_intro"] = nested_text(soup.find(id="s3b").p)
messages["s3b1"] = nested_text(soup.find(id="s3b1"))
messages["s3b2"] = nested_text(soup.find(id="s3b2"))
messages["s3b3"] = nested_text(soup.find(id="s3b3"))
# Section 4: Sui generis database rights
messages["s4_sui_generics_database_rights_titles"] = nested_text(
soup.find(id="s4")
)
messages["s4_sui_generics_database_rights_intro"] = (
soup.find(id="s4").find_next_sibling("p").string
)
s4a = nested_text(soup.find(id="s4a"))
if "nc" in license_code:
messages["s4_sui_generics_database_rights_extract_reuse_nc"] = s4a
else:
messages["s4_sui_generics_database_rights_extract_reuse"] = s4a
s4b = nested_text(soup.find(id="s4b"))
if license_code.endswith("-sa"):
messages["s4_sui_generics_database_rights_adapted_material_sa"] = s4b
else:
messages["s4_sui_generics_database_rights_adapted_material"] = s4b
messages["s4_sui_generics_database_rights_comply_s3a"] = nested_text(
soup.find(id="s4c")
)
# The next text comes after the 'ol' after s4, but isn't inside a tag itself!
parent = soup.find(id="s4").parent
s4_seen = False
take_rest = False
parts = []
for item in parent.children:
if take_rest:
if item.name == "p":
# Stop at the next paragraph
break
parts.append(str(item))
elif not s4_seen:
if isinstance(item, Tag) and item.get("id") == "s4":
s4_seen = True
continue
elif not take_rest and item.name == "ol":
# already seen s4, this is the ol, so the next child is our text
take_rest = True
messages["s4_sui_generics_database_rights_postscript"] = " ".join(parts)
# Section 5: Disclaimer
messages["s5_disclaimer_title"] = soup.find(id="s5").string
messages["s5_a"] = soup.find(id="s5a").string # bold
messages["s5_b"] = soup.find(id="s5b").string # bold
messages["s5_c"] = soup.find(id="s5c").string # not bold
# Section 6: Term and Termination
messages["s6_termination_title"] = nested_text(soup.find(id="s6"))
messages["s6_termination_applies"] = nested_text(soup.find(id="s6a"))
s6b = soup.find(id="s6b")
if s6b.p:
# most languages put the introductory text in a paragraph, making it easy
messages["s6_termination_reinstates_where"] = soup.find(
id="s6b"
).p.get_text()
else:
# if they don't, we have to pick out the text from the beginning of s6b's
# content until the beginning of the "ol" inside it.
s = ""
for child in s6b:
if child.name == "ol":
break
s += str(child)
messages["s6_termination_reinstates_where"] = s
messages["s6_termination_reinstates_automatically"] = soup.find(
id="s6b1"
).get_text()
messages["s6_termination_reinstates_express"] = soup.find(id="s6b2").get_text()
children_of_s6b = list(soup.find(id="s6b").children)
messages["s6_termination_reinstates_postscript"] = (
"".join(str(x) for x in children_of_s6b[4:7])
).strip()
messages["s6_separate_terms"] = inner_html(soup.find(id="s6c"))
messages["s6_survival"] = inner_html(soup.find(id="s6d"))
# Section 7: Other terms and conditions
messages["s7_other_terms_title"] = soup.find(id="s7").string
messages["s7_a"] = soup.find(id="s7a").string
messages["s7_b"] = soup.find(id="s7b").string
# Section 8: Interpretation
messages["s8_interpretation_title"] = soup.find(id="s8").string
for key in ["s8a", "s8b", "s8c", "s8d"]:
messages[key] = inner_html(soup.find(id=key))
validate_dictionary_is_all_text(messages)
return messages
def import_cc0_license_html(self, *, content, legalcode):
license = legalcode.license
assert license.version == "1.0", f"{license.version} is not '1.0'"
assert license.license_code == "CC0", f"{license.license_code} is not 'CC0'"
messages = {}
raw_html = content
# Parse the raw HTML to a BeautifulSoup object.
soup = BeautifulSoup(raw_html, "lxml")
deed_main_content = soup.find(id="deed-main-content")
messages["license_medium"] = inner_html(soup.find(id="deed-license").h2)
legalcode.title = messages["license_medium"]
legalcode.save()
# Big disclaimer (all caps)
messages["disclaimer"] = clean_string(nested_text(deed_main_content.blockquote))
# Statement of Purpose section: "<h3><em>Statement of Purpose</em></h3>"
messages["statement_of_purpose"] = nested_text(deed_main_content.h3)
# SOP section is formatted as paragraphs
paragraphs = deed_main_content.find_all("p")
# First 3 paragraphs in the SOP section are just text
messages["sop_p1"] = nested_text(paragraphs[0])
messages["sop_p2"] = nested_text(paragraphs[1])
messages["sop_p3"] = nested_text(paragraphs[2])
# Next paragraph is a bold term, and its definition
# <p><strong>1. Copyright and Related Rights.</strong>
# A Work... </p>
nt = name_and_text(paragraphs[3])
messages["s1_title"] = nt["name"]
messages["s1_par"] = nt["text"]
# Followed by an ordered list with 7 items
ol = paragraphs[3].find_next_sibling("ol")
for i, part in enumerate(ol.find_all("li")):
messages[f"s1_item{i}"] = nested_text(part)
# Then two more numbered paragraphs that are definitions
# <p><strong>2. Waiver.</strong> To the ...</p>
nt = name_and_text(paragraphs[4])
messages["s2_title"] = nt["name"]
messages["s2_text"] = nt["text"]
# <p><strong>3. Public License Fallback.</strong> Should...</p>
nt = name_and_text(paragraphs[5])
messages["s3_title"] = nt["name"]
messages["s3_text"] = nt["text"]
# Finally the Limitations header, no intro text, and an ol with 4 items.
# <p><strong>4. Limitations and Disclaimers.</strong></p>
s4 = paragraphs[6] # <p><strong>4. Limitations...
messages["s4_title"] = nested_text(s4)
# In English, s4 is followed by an ol with 4 items.
# In .el, s4 is followed by a <p class="tab"> with
# 3 <br/> dividing the 4 parts.
ol = s4.find_next_sibling("ol")
if ol:
for i, part in enumerate(ol.find_all("li")):
messages[f"s4_part_{i}"] = nested_text(part)
else:
p4 = s4.find_next_sibling("p", class_="tab")
text = nested_text(p4)
parts = text.split("<br />")
for i, part in enumerate(parts):
messages[f"s4_part_{i}"] = str(part)
# And that's it. The CC0 "license" is relatively short.
validate_dictionary_is_all_text(messages)
return messages
