MAINT: Simplify file identifiers generation #2003
Conversation
Codecov ReportAttention:
Additional details and impacted files@@ Coverage Diff @@
## main #2003 +/- ##
==========================================
- Coverage 94.54% 94.52% -0.02%
==========================================
Files 43 43
Lines 7549 7549
Branches 1490 1491 +1
==========================================
- Hits 7137 7136 -1
Misses 253 253
- Partials 159 160 +1 ☔ View full report in Codecov by Sentry. |
MartinThoma
changed the title
| return ByteStringObject(_rolling_checksum(stream).encode("utf8")) | ||
| def _compute_document_identifier(self) -> ByteStringObject: | ||
| md5 = hashlib.md5() | ||
| md5.update(str(time.time()).encode("utf-8")) |
There was a problem hiding this comment.
This makes document-generation non-deterministic, right?
|
What impact do the file identifiers have? Who/what makes use of them? |
MartinThoma
added
the
on-hold
|
the PDF standard says:
the identifiers are also be used for encryption. @MartinThoma so i think it's ok to make it simple. |
|
Having a deterministic way to generate PDFs is valuable to several developers. Does the current deterministic identifier generation cause any issues? |
|
first of all, it cost too much for big pdf files. |
exiledkingcc
force-pushed
the
simplify
branch
2 times, most recently
from
9092a14
to
5fd1e91
Compare
See #2003 Co-authored-by: exiledkingcc <exiledkingcc@gmail.com>
See #2003 Co-authored-by: exiledkingcc <exiledkingcc@gmail.com>
#2003 Co-authored-by: exiledkingcc <exiledkingcc@gmail.com>
#2003 Co-authored-by: exiledkingcc <exiledkingcc@gmail.com>
| @@ -1246,7 +1244,7 @@ def generate_file_identifiers(self) -> None: | |||
| id2 = self._compute_document_identifier() | |||
| else: | |||
| id1 = self._compute_document_identifier() | |||
| id2 = id1 | |||
| id2 = ByteStringObject(id1.original_bytes) | |||
There was a problem hiding this comment.
id1 is a ByteStringObject already. So .original_bytes just returns id1. Then wrapping it in ByteStringObject doesn't do anything, right?
| return ByteStringObject(_rolling_checksum(stream).encode("utf8")) | ||
| md5 = hashlib.md5() | ||
| md5.update(str(time.time()).encode("utf-8")) | ||
| md5.update(str(self.fileobj).encode("utf-8")) |
There was a problem hiding this comment.
Is self.fileobj equivalent to self._write_pdf_structure(stream)?
No description provided.