v3.3.0 Add extraction as MarkDown, include text from included XObjects

What's Changed

• Apply the font /Differences array to every byte of a text segment by @splitbrain in #453
• Normalize white-space and odd digit counts in hexadecimal strings by @splitbrain in #452
• Switch to static license badge to fix pool issues by @PrinsFrank in #473
• Rename "content" to "text" in samples to prepare for markdown extraction by @PrinsFrank in #474
• Don't parse content of inline images by @PrinsFrank in #476
• Add simple markdown document sample by @PrinsFrank in #478
• Add extracted markdown to samples by @PrinsFrank in #479
• Move text extraction code out of document parsing namespaces by @PrinsFrank in #480
• Clean up line grouping strategies by @PrinsFrank in #481
• Clean up page/document arguments as page already has access to the document by @PrinsFrank in #482
• Add sample from issue #457 by @PrinsFrank in #483
• Bump actions/checkout from 7.0.0 to 7.0.1 by @dependabot[bot] in #485
• Bump zizmorcore/zizmor-action from 0.5.7 to 0.6.0 by @dependabot[bot] in #486
• Add markdown extraction methods by @PrinsFrank in #488
• Update phpstan/phpstan requirement from 2.2.5 to 2.2.6 by @dependabot[bot] in #489
• Bump zizmorcore/zizmor-action from 0.6.0 to 0.6.1 by @dependabot[bot] in #490
• Add sample with content in xobjects by @PrinsFrank in #492
• Extract text from xObjects by @PrinsFrank in #491
• Bump zizmorcore/zizmor-action from 0.6.1 to 0.6.2 by @dependabot[bot] in #493
• Include benchmark stats as text in README by @PrinsFrank in #495
• Update comparison table with more benchmark information by @PrinsFrank in #496
• Exit recursion in xObject includes by @PrinsFrank in #498
• Fix profiling script to work without bootstrapping composer by @PrinsFrank in #499
• Track object numbers in decorated objects to detect recursion by @PrinsFrank in #500
• Use correct xObject dictionary when in xObject context by @PrinsFrank in #501
• Flush inMemoryStream to string in ContentStreamParser to reduce method call overhead by @PrinsFrank in #502
• Add operator char hashmap to reduce getOperator calls by @PrinsFrank in #503
• Don't parse contentstreams in textObjects that don't contain text when extracting text only by @PrinsFrank in #504
• Remove extra round of dictionary parsing that was then discarded by @PrinsFrank in #505
• Remove rollingCharBuffer from dictionaryParser and only keep track of two last characters by @PrinsFrank in #506
• Optimize content stream parsing by @PrinsFrank in #507
• Optimize dictionary parsing by @PrinsFrank in #508
• Remove usage of infiniteBuffer by @PrinsFrank in #509
• Remove unnecessary clones on readonly classes by @PrinsFrank in #510
• Extract bold/italic markdown information by @PrinsFrank in #511
• Extract headings in markdown by @PrinsFrank in #512
• Document supported markdown features by @PrinsFrank in #513

Full Changelog: v3.2.0...v3.3.0