Add pipeline code

2026-10-17 15:27:14 -05:00
parent 3cd5a4c658
commit 5f0e2a9768
21 changed files with 1612 additions and 0 deletions
@@ -0,0 +1,71 @@
+"""Figures: find every image in the chapters being processed and list it in figures/skipped_figures.md.
+Figures support the concept only (R7). Nothing is spoken unless a spoken sentence is approved, and none is yet,
+so every figure is listed as skipped."""
+import os
+import re
+
+import pageparts
+import pdftext
+import structure
+
+
+def toc_pictures(units, before_unit):
+    """Picture list in the table of contents: [(caption, book_page)]."""
+    lines = []
+    for u in units[:before_unit]:
+        lines += u["text"].split("\n")
+    out = []
+    for ln in lines:
+        m = re.match(r"^\s*([A-Z][A-Z .,'\-—]+?)\s*\.{3,}\s*(\d+)(?:\s*-\s*\d+)?\s*$", ln)
+        if m:
+            out.append((m.group(1).strip(), int(m.group(2))))
+    return out
+
+
+def run(book_dir, book, pdf, chapter_numbers):
+    """Writes figures/skipped_figures.md. Returns warnings."""
+    units = pdftext.units(pdf)
+    heads = structure.find_headings(units)
+    pages, _ = structure.book_pages(units, book["title"])
+    toc = toc_pictures(units, heads[0]["unit"])
+    imgs = pdftext.images(pdf)
+    rows, warnings = [], []
+    for ch in book["chapters"]:
+        if ch["number"] not in chapter_numbers:
+            continue
+        lo, hi = ch["pdf_pages"]
+        for im in imgs:
+            if not lo <= im["pdf_page"] <= hi:
+                continue
+            ui = next(i for i, u in enumerate(units) if u["pdf_page"] == im["pdf_page"] and u["half"] == im["half"])
+            caption, source = None, None
+            for u in units:
+                if u["pdf_page"] == im["pdf_page"]:
+                    for ln in u["text"].split("\n"):
+                        m = pageparts.CAPTION_RE.match(ln)
+                        if m:
+                            caption, source = m.group(2).title(), "caption printed on the page (%s)" % ln.strip()
+            if caption is None:
+                hit = [c for c, p in toc if p == pages[ui]]
+                if hit:
+                    caption, source = hit[0].title(), "table of contents picture list, page %d" % pages[ui]
+            if caption is None:
+                warnings.append(("figure-no-caption", "pdf p.%d%s" % (im["pdf_page"], im["half"] or ""), "no caption found"))
+            rows.append((ch["number"], im, pages[ui], caption, source))
+    os.makedirs(os.path.join(book_dir, "figures"), exist_ok=True)
+    out = ["# Skipped figures", "",
+           "Figures support the concept only, and nothing in a figure is read aloud unless a spoken sentence is approved.",
+           "No spoken figure sentence is approved for this book yet, so every figure below is skipped.", "",
+           "Scope: chapters %s." % ", ".join(str(n) for n in sorted(chapter_numbers)), ""]
+    for n, im, bp, caption, source in rows:
+        out += ["- Chapter %d, pdf page %d (%s half), book page %s: image %d x %d pt on the page." % (
+                    n, im["pdf_page"], {"L": "left", "R": "right", None: "whole"}[im["half"]], bp if bp and bp >= 1 else "unnumbered",
+                    im["width"], im["height"]),
+                "  Caption: %s." % (caption or "none found"),
+                "  Caption source: %s." % (source or "none"),
+                "  Reason skipped: no approved spoken sentence."]
+    if not rows:
+        out.append("No figures found in these chapters.")
+    with open(os.path.join(book_dir, "figures", "skipped_figures.md"), "w", encoding="utf-8") as f:
+        f.write("\n".join(out) + "\n")
+    return warnings