xml.etree.ElementTree.ParseError: not well-formed (invalid token): line 5863, column 12969

[jtrueblood88]

Using the newest version. I'm new to python so I'm not sure what this means.

Thank you!

File "//anaconda/bin/fbcap", line 11, in sys.exit(main()) File "//anaconda/lib/python3.5/site-packages/fbchat_archive_parser/main.py", line 66, in main app.run() File "//anaconda/lib/python3.5/site-packages/clip.py", line 652, in run self.invoke(self.parse(tokens)) File "//anaconda/lib/python3.5/site-packages/clip.py", line 634, in invoke self._main.invoke(parsed) File "//anaconda/lib/python3.5/site-packages/clip.py", line 519, in invoke self._callback({k: v for k, v in iteritems(parsed) if k not in self._subcommands}) File "//anaconda/lib/python3.5/site-packages/fbchat_archive_parser/main.py", line 27, in fbcap progress_output=sys.stdout.isatty()) File "//anaconda/lib/python3.5/site-packages/fbchat_archive_parser/parser.py", line 98, in init self.parse_content() File "//anaconda/lib/python3.5/site-packages/fbchat_archive_parser/parser.py", line 107, in __parse_content for pos, element in ET.iterparse(self.stream, events=("start", "end")): File "//anaconda/lib/python3.5/xml/etree/ElementTree.py", line 1289, in __next** for event in self._parser.read_events(): File "//anaconda/lib/python3.5/xml/etree/ElementTree.py", line 1272, in read_events raise event File "//anaconda/lib/python3.5/xml/etree/ElementTree.py", line 1230, in feed self._parser.feed(data) xml.etree.ElementTree.ParseError: not well-formed (invalid token): line 5863, column 12969

[ownaginatious]

Could you post the entire stack trace?

[jtrueblood88]

Just realized I should. See the edit above.

[ownaginatious]

Hmm, it would appear that the error is coming from the XML parsing library that's reading your messages.htm file. Are you sure the file is complete and/or not corrupted?

[jtrueblood88]

I think it's complete. I was able to use a different parser on it, but wanted to use yours as it has more options for output.

[ownaginatious]

I think I fixed the issue. Let me know if the newest version works for you (version 0.5).

[jtrueblood88]

Hmm..I think I got the same error. And it happens even if I only ask it to parse a conversation with one person.

File "//anaconda/bin/fbcap", line 11, in sys.exit(main()) File "//anaconda/lib/python3.5/site-packages/fbchat_archive_parser/main.py", line 70, in main app.run() File "//anaconda/lib/python3.5/site-packages/clip.py", line 652, in run self.invoke(self.parse(tokens)) File "//anaconda/lib/python3.5/site-packages/clip.py", line 634, in invoke self._main.invoke(parsed) File "//anaconda/lib/python3.5/site-packages/clip.py", line 519, in invoke self._callback({k: v for k, v in iteritems(parsed) if k not in self._subcommands}) File "//anaconda/lib/python3.5/site-packages/fbchat_archive_parser/main.py", line 31, in fbcap progress_output=sys.stdout.isatty()) File "//anaconda/lib/python3.5/site-packages/fbchat_archive_parser/parser.py", line 98, in init self.parse_content() File "//anaconda/lib/python3.5/site-packages/fbchat_archive_parser/parser.py", line 108, in __parse_content for pos, element in ET.iterparse(f, events=("start", "end")): File "//anaconda/lib/python3.5/xml/etree/ElementTree.py", line 1289, in __next** for event in self._parser.read_events(): File "//anaconda/lib/python3.5/xml/etree/ElementTree.py", line 1272, in read_events raise event File "//anaconda/lib/python3.5/xml/etree/ElementTree.py", line 1230, in feed self._parser.feed(data) xml.etree.ElementTree.ParseError: not well-formed (invalid token): line 5863, column 12969

[ownaginatious]

Do you happen to have the line from your messages.htm file that it's crashing on? I think it may be an encoding error on Facebook's end.

[jtrueblood88]

I think so..But what does it mean that it's column 12969? I don't see how there's that many columns..

[ownaginatious]

Well, the document coming from Facebook is one giant unformatted line of data, so perhaps that's not helpful.

I just pushed a new version (0.5.post4) to PyPI. If the issue is being caused by your document not being interpreted as UTF-8 for some reason, then that should help.

[jtrueblood88]

Okay..I'll try that. Thanks so much btw. I'll let you know if it works.

[jtrueblood88]

:( nope that didn't do it. I'll keep trying though.

[ownaginatious]

Hmm, I'm really not sure what the issue could be. Does the program run at all, or does it crash immediately?

[jtrueblood88]

It seems to run fine. When I ask it to look at a specific conversation, it skips all the other one and focuses on that one. I tried maybe having it export to csv instead of stdout, but that didn't work. I'd go in and look at the code myself but I can't even find out the environment that fbcap works from..its a bit too advanced for me.

[ownaginatious]

At this point, I think it may be because the XML parser being used is strict and technically HTML does not necessarily qualify a "strictly valid XML".

Unfortunately, the "less strict" drop-in replacement library lxml requires a lot of external dependencies that can be a pain to get installed. I'm going to try and implement something less efficient using BeautifulSoup as a fallback for situations like this. I'll respond to this ticket when it is ready.

[ownaginatious]

Okay, the tool now falls back to BeautifulSoup if something goes wrong while parsing using the iterative streaming parser. Please let me know if it works for you now (version 0.6post1).

[eric-zhu94]

Just following up on this, Error is as follows:

The streaming parser crashed due to malformed XML. Falling back to the less strict/efficient BeautifulSoup parser. This may take a while... Traceback (most recent call last): File "/Users/user/.virtualenvs/p3/lib/python3.5/site-packages/fbchat_archive_parser/parser.py", line 122, in parse_content parser=XMLParser(encoding=str('UTF-8'))): File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/xml/etree/ElementTree.py", line 1290, in __next for event in self._parser.read_events(): File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/xml/etree/ElementTree.py", line 1273, in read_events raise event File "/Library/Frameworks/Python.framework/Versions/3.5/lib/python3.5/xml/etree/ElementTree.py", line 1231, in feed self._parser.feed(data) xml.etree.ElementTree.ParseError: not well-formed (invalid token): line 3613, column 71274

During handling of the above exception, another exception occurred:

Traceback (most recent call last): File "/Users/user/.virtualenvs/p3/bin/fbcap", line 11, in sys.exit(main()) File "/Users/user/.virtualenvs/p3/lib/python3.5/site-packages/fbchat_archive_parser/main.py", line 73, in main app.run() File "/Users/user/.virtualenvs/p3/lib/python3.5/site-packages/clip.py", line 652, in run self.invoke(self.parse(tokens)) File "/Users/user/.virtualenvs/p3/lib/python3.5/site-packages/clip.py", line 634, in invoke self._main.invoke(parsed) File "/Users/user/.virtualenvs/p3/lib/python3.5/site-packages/clip.py", line 519, in invoke self._callback({k: v for k, v in iteritems(parsed) if k not in self._subcommands}) File "/Users/user/.virtualenvs/p3/lib/python3.5/site-packages/fbchat_archive_parser/main.py", line 34, in fbcap progress_output=sys.stdout.isatty()) File "/Users/user/.virtualenvs/p3/lib/python3.5/site-packages/fbchat_archive_parser/parser.py", line 98, in init self.parse_content() File "/Users/user/.virtualenvs/p3/lib/python3.5/site-packages/fbchat_archive_parser/parser.py", line 142, in __parse_content soup = BeautifulSoup(open(self.stream, 'r').read(), 'lxml') File "/Users/user/.virtualenvs/p3/lib/python3.5/site-packages/bs4/__init.py", line 156, in init** % ",".join(features)) bs4.FeatureNotFound: Couldn't find a tree builder with the features you requested: lxml. Do you need to install a parser library?

relevant line as viewed in Sublime Text:

[eric-zhu94]

Nevermind, works with lxml installed.

[ownaginatious]

@friendswithsalad thanks, looks like the lxml parser required for the fallback parser is a missing external dependency on your machine. It's a pain to install sometimes (especially on Windows), so I've switched it with html.parser, which is built into python. Let me know if it works for you now (0.6.post3)