def realHarmonic(self,dist):
qs = self.model().data(self.contextIndex, Qt.DisplayRole).toString()
word = unicode(qs) # get python string
wordLen = len(word) # save a few cycles in the test below
harmList = []
for i in range(IMC.wordCensus.size()):
word2 = unicode(IMC.wordCensus.getWord(i))
if dist >= abs(wordLen - len(word2)): # There's a chance of a match
if dist >= edit_distance(word,word2): # test it
harmList.append(word2) # one hit is on the word itself
if 1 < len(harmList) : # got at least 1 other
self.panelRef.tableModel.beginResetModel()
self.panelRef.listFilter = harmList
self.panelRef.tableModel.endResetModel()
self.panelRef.rowCountLabel.setNum(self.panelRef.proxy.rowCount())
else:
pqMsgs.infoMsg(
"There are no words in edit distance {0} edit of {1}".format(dist,word)
)
def similarWords(self):
qch = QChar(u'-')
qcp = QChar(u"'")
wordOriginal = self.model().data(self.contextIndex, Qt.DisplayRole).toString()
word = QString(wordOriginal)
word.remove(qch)
word.remove(qcp)
h1list = []
for i in range(IMC.wordCensus.size()):
word1 = IMC.wordCensus.getWord(i)
word2 = QString(word1) # force a copy!
word2.remove(qch) # otherwise this would affect the word
word2.remove(qcp) # in the census list
if 0 == word.compare(word2,Qt.CaseInsensitive):
h1list.append(unicode(word1)) # one hit on word itself
if 1 < len(h1list): # got at least 1 other
self.panelRef.tableModel.beginResetModel()
self.panelRef.listFilter = h1list
self.panelRef.tableModel.endResetModel()
self.panelRef.rowCountLabel.setNum(self.panelRef.proxy.rowCount())
else:
pqMsgs.infoMsg("There are no words similar to {0}".format(unicode(wordOriginal)))
def checkForUpdates():
# Step one, read the code display for the master
page = slurp('https://github.com/tallforasmurf/PPQT')
if page is None:
pqMsgs.warningMsg('Unable to contact Github',
'Perhaps try again later?')
return
# the embedded URLs usable.
hit = reGetMasterSHA.search(page)
if hit is None:
pqMsgs.warningMsg('Github page format not as expected',
'Probable bug, update not available')
return
masterSHA = hit.group(1)
page = re.sub(masterSHA, 'master', page)
# Step three, make a list of all the "blobs" mentioned, which is just
# all the committed modules, pqXXX.py and pqHelp.html. Ignoring the
# extras -- the extras folder appears as a "tree" item and we could
# follow it and list all the blobs extras/* but we are not.
blobs = reBlobLine.findall(page)
# blobs is a list of 4-tuples, the (/1, /2, /3, /4) from reBlobLine above
# Now make a dict with 3-item values,
# { modname:[ None, masterhash, masterURL] } where None will be filled
# in with the local hash value shortly.
blobTab = {}
for blob in blobs:
blobURL = u'https://raw.github.com' + blob[0] + u'/' + blob[1]
blobName = blob[3]
blobHash = blob[2]
blobTab[blobName] = [None, blobHash, blobURL]
# Look for the file modulehash.txt and if it exists, use it to fill
# in the local hashes it lists. We do not distribute so it doesn't
# show up in the github listing. We want the local hash to reflect
# the actual local files, which might have been diddled locally.
mhPath = os.path.join(IMC.appBasePath, 'modulehashes.txt')
try:
mhFile = open(mhPath, 'r')
for line in mhFile:
[modname, localhash] = line.split()
if modname in blobTab:
blobTab[modname][0] = localhash
except:
# presumably the file doesn't exist
pass
# Run through the blobTab and try to get hash values for any
# modules that don't have a local hash yet (because modulehashes
# didn't exist -- or didn't list them because they're new to us).
for modName in blobTab:
if blobTab[modName][0] is None:
modPath = os.path.join(IMC.appBasePath, modName)
try:
modFile = open(modPath, 'r')
modText = modFile.read()
hasher = initializedSHA1.copy()
hasher.update(str(len(modText)))
hasher.update(b'\000')
hasher.update(modText)
blobTab[modName][0] = hasher.hexdigest()
except:
# presumably modname doesn't exist (new module?)
pass
# Run through the blobTab and make a new table, updaTab, listing
# the members where localhash differs from master hash.
updaTab = {}
for modName in blobTab:
if blobTab[modName][0] != blobTab[modName][1]:
updaTab[modName] = blobTab[modName]
# If there are no names left in the updaTab, the app is up to date!
if len(updaTab) == 0:
pqMsgs.infoMsg('PPQT is up to date.')
return
# There are one or more modules needing updating. Ask the user
# if we should proceed.
ans = pqMsgs.okCancelMsg(
'{0} module(s) can be updated.'.format(len(updaTab)),
'Shall we proceed?')
if not ans:
pqMsgs.infoMsg('PPQT remains unchanged.')
return
# User said OK to do it. Read the text of the updated modules from
# github and save it in the updaTab.
for modName in updaTab:
page = slurp(updaTab[modName][2])
if page is None:
pqMsgs.warningMsg('Some problem reading update modules',
'PPQT is unchanged.')
return
updaTab[modName].append(page)
# All update texts read correctly. Now open each for writing,
# appending the file object to the updaTab entry.
for modName in updaTab:
try:
modPath = os.path.join(IMC.appBasePath, modName)
modFile = open(modPath, 'w')
updaTab[modName].append(modFile)
except Exception:
pqMsgs.warningMsg('Updated modules are not writable'
'PPQT is unchanged.')
return
# All files open for writing and text is ready. Write them.
for modName in updaTab:
try:
modFile = updaTab[modName][4]
modFile.write(updaTab[modName][3])
modFile.flush()
os.fsync(modFile.fileno())
modFile.close()
except Exception as e:
# This is the bad case: some amount of writing done but not
# all of it complete. PPQT is in an inconsistent state.
pqMsgs.warningMsg(
'Error writing updated module(s)!',
'PPQT is in an inconsistent state\nDownload a complete new copy'
)
return
# All updates complete. Record local hashes in modulehashes.txt
try:
mhFile = open(mhPath, 'w')
for modName in blobTab:
mhFile.write(modName + ' ' + blobTab[modName][1] + '\n')
mhFile.close()
except Exception as e:
pass
pqMsgs.infoMsg('Updates applied.',
'Changes take effect when PPQT is restarted.')
return
def checkForUpdates():
# Step one, read the code display for the master
page = slurp('https://github.com/tallforasmurf/PPQT')
if page is None:
pqMsgs.warningMsg('Unable to contact Github',
'Perhaps try again later?')
return
# the embedded URLs usable.
hit = reGetMasterSHA.search(page)
if hit is None:
pqMsgs.warningMsg('Github page format not as expected',
'Probable bug, update not available')
return
masterSHA = hit.group(1)
page = re.sub(masterSHA,'master',page)
# Step three, make a list of all the "blobs" mentioned, which is just
# all the committed modules, pqXXX.py and pqHelp.html. Ignoring the
# extras -- the extras folder appears as a "tree" item and we could
# follow it and list all the blobs extras/* but we are not.
blobs = reBlobLine.findall(page)
# blobs is a list of 4-tuples, the (/1, /2, /3, /4) from reBlobLine above
# Now make a dict with 3-item values,
# { modname:[ None, masterhash, masterURL] } where None will be filled
# in with the local hash value shortly.
blobTab = {}
for blob in blobs:
blobURL = u'https://raw.github.com' + blob[0] + u'/' + blob[1]
blobName = blob[3]
blobHash = blob[2]
blobTab[blobName] =[None,blobHash,blobURL]
# Look for the file modulehash.txt and if it exists, use it to fill
# in the local hashes it lists. We do not distribute so it doesn't
# show up in the github listing. We want the local hash to reflect
# the actual local files, which might have been diddled locally.
mhPath = os.path.join(IMC.appBasePath,'modulehashes.txt')
try:
mhFile = open(mhPath,'r')
for line in mhFile:
[modname, localhash] = line.split()
if modname in blobTab:
blobTab[modname][0] = localhash
except:
# presumably the file doesn't exist
pass
# Run through the blobTab and try to get hash values for any
# modules that don't have a local hash yet (because modulehashes
# didn't exist -- or didn't list them because they're new to us).
for modName in blobTab:
if blobTab[modName][0] is None:
modPath = os.path.join(IMC.appBasePath,modName)
try:
modFile = open(modPath,'r')
modText = modFile.read()
hasher = initializedSHA1.copy()
hasher.update(str(len(modText)))
hasher.update(b'\000')
hasher.update(modText)
blobTab[modName][0] = hasher.hexdigest()
except:
# presumably modname doesn't exist (new module?)
pass
# Run through the blobTab and make a new table, updaTab, listing
# the members where localhash differs from master hash.
updaTab = {}
for modName in blobTab:
if blobTab[modName][0] != blobTab[modName][1] :
updaTab[modName] = blobTab[modName]
# If there are no names left in the updaTab, the app is up to date!
if len(updaTab) == 0:
pqMsgs.infoMsg('PPQT is up to date.')
return
# There are one or more modules needing updating. Ask the user
# if we should proceed.
ans = pqMsgs.okCancelMsg('{0} module(s) can be updated.'.format(len(updaTab)),
'Shall we proceed?')
if not ans:
pqMsgs.infoMsg('PPQT remains unchanged.')
return
# User said OK to do it. Read the text of the updated modules from
# github and save it in the updaTab.
for modName in updaTab:
page = slurp(updaTab[modName][2])
if page is None:
pqMsgs.warningMsg('Some problem reading update modules',
'PPQT is unchanged.')
return
updaTab[modName].append(page)
# All update texts read correctly. Now open each for writing,
# appending the file object to the updaTab entry.
for modName in updaTab:
try:
modPath = os.path.join(IMC.appBasePath,modName)
modFile = open(modPath, 'w')
updaTab[modName].append(modFile)
except Exception:
pqMsgs.warningMsg('Updated modules are not writable'
'PPQT is unchanged.')
return
# All files open for writing and text is ready. Write them.
for modName in updaTab:
try:
modFile = updaTab[modName][4]
modFile.write(updaTab[modName][3])
modFile.flush()
os.fsync(modFile.fileno())
modFile.close()
except Exception as e:
# This is the bad case: some amount of writing done but not
# all of it complete. PPQT is in an inconsistent state.
pqMsgs.warningMsg('Error writing updated module(s)!',
'PPQT is in an inconsistent state\nDownload a complete new copy')
return
# All updates complete. Record local hashes in modulehashes.txt
try:
mhFile = open(mhPath,'w')
for modName in blobTab:
mhFile.write(modName + ' ' + blobTab[modName][1] + '\n')
mhFile.close()
except Exception as e:
pass
pqMsgs.infoMsg('Updates applied.',
'Changes take effect when PPQT is restarted.')
return
def loadMetadata(self,metaStream):
sectionRE = QRegExp( u"\{\{(" + '|'.join (
['PAGETABLE','CHARCENSUS','WORDCENSUS','BOOKMARKS',
'NOTES','GOODWORDS','BADWORDS','CURSOR','VERSION',
'STALECENSUS','NEEDSPELLCHECK','ENCODING', 'DOCHASH', 'MAINDICT'] ) \
+ u")(.*)\}\}",
Qt.CaseSensitive)
metaVersion = 0 # base version
while not metaStream.atEnd() :
qline = metaStream.readLine().trimmed()
if qline.isEmpty() : continue # allow blank lines between sections
if sectionRE.exactMatch(qline) : # section start
section = sectionRE.cap(1)
argument = unicode(sectionRE.cap(2).trimmed())
endsec = QString(u"{{/" + section + u"}}")
if section == u"VERSION":
if len(argument) != 0 :
metaVersion = int(argument)
continue # no more data after {{VERSION x }}
elif section == u"STALECENSUS" :
if argument == u"TRUE" :
IMC.staleCensus = IMC.staleCensusLoaded
continue # no more data after {{STALECENSUS x}}
elif section == u"NEEDSPELLCHECK" :
if argument == u"TRUE" :
IMC.needSpellCheck = True
continue # no more data after {{NEEDSPELLCHECK x}}
elif section == u"ENCODING" :
IMC.bookSaveEncoding = QString(argument)
continue
elif section == u"MAINDICT" :
IMC.bookMainDict = QString(argument)
continue
elif section == u"DOCHASH" :
IMC.metaHash = argument
continue
elif section == u"PAGETABLE":
qline = metaStream.readLine()
while (not qline.startsWith(endsec)) and (not qline.isEmpty()):
IMC.pageTable.metaStringIn(qline)
qline = metaStream.readLine()
continue
elif section == u"CHARCENSUS":
qline = metaStream.readLine()
while (not qline.startsWith(endsec)) and (not qline.isEmpty()):
# can't just .split the char census, the first
# char is the char being counted and it can be a space.
str = unicode(qline)
parts = str[2:].split(' ')
IMC.charCensus.append(QString(str[0]),int(parts[0]),int(parts[1]))
qline = metaStream.readLine()
continue
elif section == u"WORDCENSUS":
qline = metaStream.readLine()
while (not qline.startsWith(endsec)) and (not qline.isEmpty()):
parts = unicode(qline).split(' ')
IMC.wordCensus.append(QString(parts[0]),int(parts[1]),int(parts[2]))
qline = metaStream.readLine()
continue
elif section == u"BOOKMARKS":
qline = metaStream.readLine()
while (not qline.startsWith(endsec)) and (not qline.isEmpty()):
parts = unicode(qline).split(' ')
tc = QTextCursor(self.document() )
tc.setPosition(int(parts[1]))
if len(parts) == 3 : # early versions didn't save anchor
tc.movePosition(int(parts[2]),QTextCursor.KeepAnchor)
self.bookMarkList[int(parts[0])] = tc
qline = metaStream.readLine()
continue
elif section == u"NOTES":
e = IMC.notesEditor
e.setUndoRedoEnabled(False)
qline = metaStream.readLine()
while (not qline.startsWith(endsec)) and not metaStream.atEnd():
if qline.startsWith(u"\xfffd"): # escaped {{
qline.remove(0,1)
e.appendPlainText(qline)
qline = metaStream.readLine()
e.setUndoRedoEnabled(True)
continue
elif section == u"GOODWORDS" :
# not going to bother checking for endsec return,
# if it isn't that then we will shortly fail anyway
w = IMC.goodWordList.load(metaStream,endsec)
continue
elif section == u"BADWORDS" :
w = IMC.badWordList.load(metaStream,endsec)
continue
elif section == u"CURSOR" : # restore selection as of save
p1p2 = argument.split(' ')
tc = QTextCursor(self.document())
tc.setPosition(int(p1p2[0]),QTextCursor.MoveAnchor)
tc.setPosition(int(p1p2[1]),QTextCursor.KeepAnchor)
self.setTextCursor(tc)
else:
# this can't happen; section is text captured by the RE
# and we have accounted for all possibilities
raise AssertionError, "impossible metadata"
else: # Non-blank line that doesn't match sectionRE?
pqMsgs.infoMsg(
"Unexpected line in metadata: {0}".format(pqMsgs.trunc(qline,20)),
"Metadata may be incomplete, suggest quit")
break
def loadMetadata(self, metaStream):
sectionRE = QRegExp( u"\{\{(" + '|'.join (
['PAGETABLE','CHARCENSUS','WORDCENSUS','BOOKMARKS',
'NOTES','GOODWORDS','BADWORDS','CURSOR','VERSION',
'STALECENSUS','NEEDSPELLCHECK','ENCODING', 'DOCHASH', 'MAINDICT'] ) \
+ u")(.*)\}\}",
Qt.CaseSensitive)
metaVersion = 0 # base version
while not metaStream.atEnd():
qline = metaStream.readLine().trimmed()
if qline.isEmpty(): continue # allow blank lines between sections
if sectionRE.exactMatch(qline): # section start
section = sectionRE.cap(1)
argument = unicode(sectionRE.cap(2).trimmed())
endsec = QString(u"{{/" + section + u"}}")
if section == u"VERSION":
if len(argument) != 0:
metaVersion = int(argument)
continue # no more data after {{VERSION x }}
elif section == u"STALECENSUS":
if argument == u"TRUE":
IMC.staleCensus = IMC.staleCensusLoaded
continue # no more data after {{STALECENSUS x}}
elif section == u"NEEDSPELLCHECK":
if argument == u"TRUE":
IMC.needSpellCheck = True
continue # no more data after {{NEEDSPELLCHECK x}}
elif section == u"ENCODING":
IMC.bookSaveEncoding = QString(argument)
continue
elif section == u"MAINDICT":
IMC.bookMainDict = QString(argument)
continue
elif section == u"DOCHASH":
IMC.metaHash = argument
continue
elif section == u"PAGETABLE":
qline = metaStream.readLine()
while (not qline.startsWith(endsec)) and (
not qline.isEmpty()):
IMC.pageTable.metaStringIn(qline)
qline = metaStream.readLine()
continue
elif section == u"CHARCENSUS":
qline = metaStream.readLine()
while (not qline.startsWith(endsec)) and (
not qline.isEmpty()):
# can't just .split the char census, the first
# char is the char being counted and it can be a space.
str = unicode(qline)
parts = str[2:].split(' ')
IMC.charCensus.append(QString(str[0]), int(parts[0]),
int(parts[1]))
qline = metaStream.readLine()
continue
elif section == u"WORDCENSUS":
qline = metaStream.readLine()
while (not qline.startsWith(endsec)) and (
not qline.isEmpty()):
parts = unicode(qline).split(' ')
IMC.wordCensus.append(QString(parts[0]), int(parts[1]),
int(parts[2]))
qline = metaStream.readLine()
continue
elif section == u"BOOKMARKS":
qline = metaStream.readLine()
while (not qline.startsWith(endsec)) and (
not qline.isEmpty()):
parts = unicode(qline).split(' ')
tc = QTextCursor(self.document())
tc.setPosition(int(parts[1]))
if len(parts
) == 3: # early versions didn't save anchor
tc.movePosition(int(parts[2]),
QTextCursor.KeepAnchor)
self.bookMarkList[int(parts[0])] = tc
qline = metaStream.readLine()
continue
elif section == u"NOTES":
e = IMC.notesEditor
e.setUndoRedoEnabled(False)
qline = metaStream.readLine()
while (not qline.startsWith(endsec)
) and not metaStream.atEnd():
if qline.startsWith(u"\xfffd"): # escaped {{
qline.remove(0, 1)
e.appendPlainText(qline)
qline = metaStream.readLine()
e.setUndoRedoEnabled(True)
continue
elif section == u"GOODWORDS":
# not going to bother checking for endsec return,
# if it isn't that then we will shortly fail anyway
w = IMC.goodWordList.load(metaStream, endsec)
continue
elif section == u"BADWORDS":
w = IMC.badWordList.load(metaStream, endsec)
continue
elif section == u"CURSOR": # restore selection as of save
p1p2 = argument.split(' ')
tc = QTextCursor(self.document())
tc.setPosition(int(p1p2[0]), QTextCursor.MoveAnchor)
tc.setPosition(int(p1p2[1]), QTextCursor.KeepAnchor)
self.setTextCursor(tc)
else:
# this can't happen; section is text captured by the RE
# and we have accounted for all possibilities
raise AssertionError, "impossible metadata"
else: # Non-blank line that doesn't match sectionRE?
pqMsgs.infoMsg(
"Unexpected line in metadata: {0}".format(
pqMsgs.trunc(qline, 20)),
"Metadata may be incomplete, suggest quit")
break