v2.0.0 Alpha 1

What's Changed

Positional text extraction

PDF text objects have a x and y coordinates besides other translations in their position matrix. After a lot of hours this is now extracted and parsed, meaning that text that appears across multiple lines will no longer appear as a single line in the text output!

Other changes

• More strictly parse dictionary arrays by @PrinsFrank in #53
• Add new PDF2.0 tabs nameValues by @PrinsFrank in #54
• Fix issue where nested arrays in dictionaries were closed too early by @PrinsFrank in #56
• Fix comment state in dictionary parsing by @PrinsFrank in #58
• Fix nesting issues in text operator parsing by @PrinsFrank in #60
• Correctly parse dictionary array values where items are seperated over several lines by @PrinsFrank in #62
• Add missing public key security handlers to FilterNameValue by @PrinsFrank in #65
• Add support for multiple codespaceranges by @PrinsFrank in #67
• Remove extra decoding in uncompressed object by @PrinsFrank in #68
• Allow dashes in font names by @PrinsFrank in #70
• Fix issues where font operator contains font names by @PrinsFrank in #72
• Rename TextObjects and collection to contentStream to better reflect expected content by @PrinsFrank in #73
• Tc operator sets char space not char size by @PrinsFrank in #74
• Keep track of content stream commands outside text objects by @PrinsFrank in #75
• Remove dependency to internal phpunit method and ignore non-testing tests by @PrinsFrank in #76
• Organize content stream classes by @PrinsFrank in #77
• Organize text operators by @PrinsFrank in #78
• Parse textObjects to intermediate PositionedTextElement by @PrinsFrank in #79
• Implement remaining content stream operators by @PrinsFrank in #81
• Fix false positives on content stream commands by @PrinsFrank in #82
• Implement a transformation state stack to fix text lines appearing out of order by @PrinsFrank in #83
• Fix a typo in CONTRIBUTING by @szepeviktor in #84
• Fix content stream unit test after matrix multiplication fix resulting in correct x and y offsets by @PrinsFrank in #85
• Add test for compressed byte offsets by @PrinsFrank in #87
• Parse CIDFontWidths in dictionary values by @PrinsFrank in #89
• Implement space insertion based on text width by @PrinsFrank in #90
• Update samples dependency and increase space insertion threshold by @PrinsFrank in #91
• Don't try to parse EMCs and other content stream content as dictionaries by @PrinsFrank in #92
• Don't continue matching operators when in dictionary key or escaped string by @PrinsFrank in #94
• Implement encryption detection by @PrinsFrank in #97
• Use object length instead of searching for end marker when set in dictionary by @PrinsFrank in #95
• Add missing tests for inMemoryStream by @PrinsFrank in #98

New Contributors

• @szepeviktor made their first contribution in #84

Full Changelog: v1.1.0...v2.0.0-alpha.1