blob: 92da6977e932ece5d8bbee20c9d0af44691b8faa (
plain) (
blame)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
|
#!/usr/bin/env bash
# pdf-txt — extract clean text from OCR'd PDFs, recursively
#
# Usage: pdf-txt <path> [<path>...]
#
# <path> can be a .pdf file or a directory (walked recursively).
# For each PDF with a text layer, writes a sibling .txt file using
# pdftotext -layout, with "=== Page N ===" markers inserted.
# Skips PDFs without a text layer.
# Skips PDFs whose .txt is newer than the source.
set -eu
die() { printf "ERROR: %s\n" "$*" >&2; exit 1; }
need() { command -v "$1" >/dev/null 2>&1 || die "Missing: $1 (apt install $2)"; }
need pdftotext poppler-utils
need find findutils
need awk gawk
processed=0; skipped_no_ocr=0; skipped_existing=0; failed=0
extract_one() {
local pdf="$1"
local txt="${pdf%.[Pp][Dd][Ff]}.txt"
if [ -f "$txt" ] && [ "$txt" -nt "$pdf" ]; then
printf " up-to-date %s\n" "$pdf"
skipped_existing=$((skipped_existing + 1))
return 0
fi
# OCR/text-layer detection: sample first 3 pages
local tmp; tmp=$(mktemp)
pdftotext -l 3 "$pdf" "$tmp" 2>/dev/null || true
local nchars; nchars=$(tr -d '[:space:]' < "$tmp" | wc -c)
rm -f "$tmp"
if [ "$nchars" -lt 50 ]; then
printf " no OCR %s\n" "$pdf"
skipped_no_ocr=$((skipped_no_ocr + 1))
return 0
fi
if pdftotext -layout "$pdf" - 2>/dev/null | awk '
BEGIN { page = 1; printf "=== Page %d ===\n\n", page }
/\014/ { page++; printf "\n=== Page %d ===\n\n", page; next }
{ print }
' > "$txt"; then
printf " extracted %s\n" "$txt"
processed=$((processed + 1))
else
rm -f "$txt"
printf " FAILED %s\n" "$pdf" >&2
failed=$((failed + 1))
fi
}
walk() {
local target="$1"
if [ -f "$target" ]; then
case "$target" in
*.pdf|*.PDF) extract_one "$target" ;;
*) printf " not a PDF %s\n" "$target" >&2 ;;
esac
elif [ -d "$target" ]; then
while IFS= read -r -d '' pdf; do
extract_one "$pdf"
done < <(find "$target" -type f -iname '*.pdf' -print0 | sort -z)
else
printf " not found %s\n" "$target" >&2
fi
}
[ $# -ge 1 ] || die "Usage: pdf-txt <path> [<path>...]"
for arg in "$@"; do walk "$arg"; done
printf "\nDone. processed=%d no_ocr=%d up_to_date=%d failed=%d\n" \
"$processed" "$skipped_no_ocr" "$skipped_existing" "$failed"
|
