Repository navigation
Conversation
…Explorer Addresses programminghistorian#3860 with targeted guidance updates that keep the lesson's approach: - Add an update notice that lists the changes and the tool versions used to check the commands. - Replace the Windows 10 / Ubuntu 18.04 WSL steps with `wsl --install`, the obsolete Homebrew installer with the current one, and the Ubuntu 18.04 instructions with current Ubuntu LTS. Install only the English Tesseract data and explain how to add other languages. - Never overwrite source PDFs: OCRmyPDF, pdftotext and pdfimages now write to separate `ocr/`, `text/` and `images/` folders with predictable file names. - Use `--skip-text` for batch OCR, since OCRmyPDF aborts on pages that already contain text by default. Drop `--clean` from the default commands, explain that it needs unpaper, and mention `--redo-ocr` and `--force-ocr` for bad text layers. - Restrict the grep examples to the `text/` folder. - Use ImageMagick 7's `magick` command and explain the `convert` fallback and the img2pdf workaround for the PDF security policy. - Replace the ILO download one-liner with a readable loop that reports the file count, fails loudly, can resume, and points to Zenodo. - Replace DARIAH Topics Explorer with Simple Topic Modeling (https://maehr.github.io/simple-topic-modeling/): NMF with TF-IDF as the default, LDA as an option, stop words, random seed, diagnostics, and repeatable runs through config.json. Remove the DARIAH screenshot. - Update the WSL, grep and topic-model references. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_018k44PqSnqQ4gfSwQHwq2E9
There was a problem hiding this comment.
Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.
An organization owner can choose the repository organization as the billing source for reviews requested by members with a paid Copilot license. Learn about Copilot code review billing.
|
@anisa-hawes, thanks for reviewing this PR. As you can see, I used Claude (Opus 5.5/High) for assistance. I hope this is okay with you and with the Programming Historian's AI rules. |
Check the ILO corpus size before downloads. Preserve complete outputs and report processing failures. Correct dependency and cleaning guidance. Remove the unsupported version claim.
|
Hello Moritz @maehr, To reiterate Laura's comment in Issue 3887, I’m closing this PR. If you’d like to propose short information or guidance panels, please add them to the issue thread and indicate where they should appear in the lesson. I’ll then take care of implementing the changes in a new PR. As Laura @lachapot explained, all direct updates to our site are managed by our team. Thank you. |
Related to #3860.
This revision updates installation guidance and makes PDF downloads, OCR and text extraction resumable. It preserves the PDF → OCR → text extraction → corpus analysis workflow and replaces DARIAH Topics Explorer with Simple Topic Modeling.
Changes
--skip-textand--deskewin batch OCR. Explain--redo-ocr,--force-ocr, optional cleaning and changes to final page images.img2pdfinstallation.config.json. Preserve the existing DOI and configuration version.Validation
bash -n.--redo-ocrand--force-ocrreplaced deliberately incorrect hidden text.git diff --check. Inspected a Pandoc HTML preview and verified that rendered code blocks preserve the commands.Remaining validation limits
bundle exec jekyll buildfailed locally because Bundler 2.4.22 is unavailable. Chrome was unavailable for visual browser review. The Pandoc preview does not replace a Jekyll build.