def text_history(ref, version, lang):
"""
Return a complete list of changes to a segment of text (identified by ref/version/lang)
"""
ref = texts.norm_ref(ref)
refRe = '^%s$|^%s:' % (ref, ref)
changes = texts.db.history.find({"ref": {"$regex": refRe}, "version": version, "language": lang}).sort([['revision', -1]])
history = []
for i in range(changes.count()):
rev = changes[i]
log = {
"ref": rev["ref"],
"revision": rev["revision"],
"date": rev["date"],
"user": rev["user"],
"rev_type": rev["rev_type"],
"method": rev.get("method", "Site"),
"diff_html": rev["diff_html"],
"text": text_at_revision(ref, version, lang, rev["revision"])
}
history.append(log)
"""
# create a fake revision 0 for initial work that was unrecorded
rev0 = {
"revision": 0,
"date": "Date Unknown",
"user": "******",
"rev_type": "add text",
"diff_html": text_at_revision(ref, version, lang, 0)
}
history.append(rev0)
"""
return history
def text_history(ref, version, lang):
"""
Return a complete list of changes to a segment of text (identified by ref/version/lang)
"""
ref = texts.norm_ref(ref)
changes = texts.db.history.find({
"ref": ref,
"version": version,
"language": lang
}).sort([['revision', -1]])
history = []
for i in range(changes.count()):
rev = changes[i]
log = {
"revision": rev["revision"],
"date": rev["date"],
"user": rev["user"],
"rev_type": rev["rev_type"],
"diff_html": rev["diff_html"],
"text": text_at_revision(ref, version, lang, rev["revision"])
}
history.append(log)
# create a fake revision 0 for initial work that was unrecorded
rev0 = {
"revision": 0,
"date": "Date Unknown",
"user": "******",
"rev_type": "add text",
"diff_html": text_at_revision(ref, version, lang, 0)
}
history.append(rev0)
return history
def record_text_change(ref, version, lang, text, user, **kwargs):
"""
Record a change to a text (ref/version/lang) by user.
"""
# unpack text into smaller segments if necessary (e.g. chapter -> verse)
if isinstance(text, list):
for i in reversed(range(len(text))):
n = i + 1
record_text_change("%s.%d" % (ref, n), version, lang, text[i], user, **kwargs)
return
# get the current state of the text in question
current = texts.get_text(ref, context=0, commentary=False, version=version, lang=lang)
if "error" in current and current["error"].startswith("No text found"):
current = ""
elif "error" in current:
return current
elif lang == "en" and current["text"]:
current = current["text"]
elif lang == "he" and current["he"]:
current = current["he"]
else:
current = ""
# Don't record anything if there's no change.
if not text:
text = ""
if text == current:
return
# create a patch that turns the new version back into the old
backwards_diff = dmp.diff_main(text, current)
patch = dmp.patch_toText(dmp.patch_make(backwards_diff))
# get html displaying edits in this change.
forwards_diff = dmp.diff_main(current, text)
dmp.diff_cleanupSemantic(forwards_diff)
diff_html = dmp.diff_prettyHtml(forwards_diff)
# give this revision a new revision number
revision = next_revision_num()
log = {
"ref": texts.norm_ref(ref),
"version": version,
"language": lang,
"diff_html": diff_html,
"revert_patch": patch,
"user": user,
"date": datetime.now(),
"revision": revision,
"message": kwargs.get("message", ""),
"rev_type": kwargs.get("type", None) or "edit text" if len(current) else "add text",
"method": kwargs.get("method", "Site")
}
db.history.save(log)
def record_text_change(ref, version, lang, text, user, **kwargs):
"""
Record a change to a text (ref/version/lang) by user.
"""
# unpack text into smaller segments if necessary (e.g. chapter -> verse)
if isinstance(text, list):
for i in reversed(range(len(text))):
n = i + 1
record_text_change("%s.%d" % (ref, n), version, lang, text[i], user, **kwargs)
return
# get the current state of the text in question
current = texts.get_text(ref, context=0, commentary=False, version=version, lang=lang)
if "error" in current and current["error"].startswith("No text found"):
current = ""
elif "error" in current:
return current
elif lang == "en" and current["text"]:
current = current["text"]
elif lang == "he" and current["he"]:
current = current["he"]
else:
current = ""
# Don't record anything if there's no change.
if not text:
text = ""
if text == current:
return
# create a patch that turns the new version back into the old
backwards_diff = dmp.diff_main(text, current)
patch = dmp.patch_toText(dmp.patch_make(backwards_diff))
# get html displaying edits in this change.
forwards_diff = dmp.diff_main(current, text)
dmp.diff_cleanupSemantic(forwards_diff)
diff_html = dmp.diff_prettyHtml(forwards_diff)
# give this revision a new revision number
revision = next_revision_num()
log = {
"ref": texts.norm_ref(ref),
"version": version,
"language": lang,
"diff_html": diff_html,
"revert_patch": patch,
"user": user,
"date": datetime.now(),
"revision": revision,
"message": kwargs.get("message", ""),
"rev_type": kwargs.get("type", None) or "edit text" if len(current) else "add text",
"method": kwargs.get("method", "Site")
}
texts.db.history.save(log)
def text_history(ref, version, lang, filter_type=None):
"""
Return a complete list of changes to a segment of text (identified by ref/version/lang)
"""
ref = texts.norm_ref(ref)
refRe = '^%s$|^%s:' % (ref, ref)
query = {"ref": {"$regex": refRe}, "version": version, "language": lang}
query.update(filter_type_to_query(filter_type))
return get_activity(query, page_size=0, page=1, filter_type=filter_type)
def get_last_edit(ref, lang, version):
"""
Returns the last edit or addition to ref/lang/version
"""
ref = texts.norm_ref(ref)
refRe = '^%s$|^%s:' % (ref, ref)
query = {"ref": {"$regex": refRe}, "language": lang, "version": version,
"rev_type": {"$in": ["edit text", "add text", "revert text"]}}
edit = texts.db.history.find(query).sort([["date", -1]]).limit(1)
if edit.count():
return edit[0]
return None
def get_reviews(ref, lang, version):
"""
Returns a list of reviews pertaining to ref/lang/version
"""
reviews = []
ref = texts.norm_ref(ref)
refRe = '^%s$|^%s:' % (ref, ref)
cursor = texts.db.history.find({"ref": {"$regex": refRe}, "language": lang, "version": version, "rev_type": "review"}).sort([["date", -1]])
for r in cursor:
r["_id"] = str(r["_id"])
r["userLink"] = user_link(r["user"])
reviews.append(r)
return reviews
def index_text(ref, version=None, lang=None):
"""
Index the text designated by ref.
If no version and lang are given, this functon will be called for each availble version.
Currently assumes ref is at section level.
"""
ref = texts.norm_ref(unicode(ref))
# Recall this function for each specific text version, if non provided
if not (version and lang):
for v in texts.get_version_list(ref):
index_text(ref, version=v["versionTitle"], lang=v["language"])
return
# Index each segment of this document individually
pRef = texts.parse_ref(ref)
if len(pRef["sections"]) < len(pRef["sectionNames"]):
text = texts.get_text(ref,
context=0,
commentary=False,
version=version,
lang=lang)
if "error" in text:
print text["error"]
else:
for i in range(max(len(text["text"]), len(text["he"]))):
index_text("%s:%d" % (ref, i + 1))
# Don't try to index docs with depth 3
if len(pRef["sections"]) < len(pRef["sectionNames"]) - 1:
return
# Index this document as a whole
doc = make_text_index_document(ref, version, lang)
if doc:
try:
es.index(doc, 'sefaria', 'text',
make_text_doc_id(ref, version, lang))
global doc_count
doc_count += 1
except Exception, e:
print "Error indexing %s / %s / %s" % (ref, version, lang)
print e
def index_text(ref, version=None, lang=None):
"""
Index the text designated by ref.
If no version and lang are given, this functon will be called for each availble version.
Currently assumes ref is at section level.
"""
ref = texts.norm_ref(unicode(ref))
# Recall this function for each specific text version, if non provided
if not (version and lang):
for v in texts.get_version_list(ref):
index_text(ref, version=v["versionTitle"], lang=v["language"])
return
# Index each segment of this document individually
pRef = texts.parse_ref(ref)
if len(pRef["sections"]) < len(pRef["sectionNames"]):
text = texts.get_text(ref, context=0, commentary=False, version=version, lang=lang)
if "error" in text:
print text["error"]
else:
for i in range(max(len(text["text"]), len(text["he"]))):
index_text("%s:%d" % (ref, i+1))
# Don't try to index docs with depth 3
if len(pRef["sections"]) < len(pRef["sectionNames"]) - 1:
return
# Index this document as a whole
doc = make_text_index_document(ref, version, lang)
if doc:
try:
global doc_count
if doc_count % 5000 == 0:
print "[%d] Indexing %s / %s / %s" % (doc_count, ref, version, lang)
es.index('sefaria', 'text', doc, make_text_doc_id(ref, version, lang))
doc_count += 1
except Exception, e:
print "ERROR indexing %s / %s / %s" % (ref, version, lang)
pprint(e)
