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
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
//! Debug extraction - analyze text spans and gaps for quality issues
//!
//! Usage:
//! cargo run --bin debug_extraction -- pdf_file.pdf
//! RUST_LOG=debug cargo run --bin debug_extraction -- pdf_file.pdf
use pdf_oxide::document::PdfDocument;
use std::env;
fn main() -> Result<(), Box<dyn std::error::Error>> {
env_logger::init();
let args: Vec<String> = env::args().collect();
if args.len() < 2 {
eprintln!("Usage: {} <pdf_file>", args[0]);
std::process::exit(1);
}
let pdf_path = &args[1];
println!("\n=== PDF EXTRACTION DEBUG ===");
println!("File: {}\n", pdf_path);
let doc = PdfDocument::open(pdf_path)?;
let page_count = doc.page_count()?;
// Analyze spans for first page (or specified page)
let page_num = args
.get(2)
.and_then(|s| s.parse::<usize>().ok())
.unwrap_or(0);
if page_num >= page_count {
eprintln!("Page {} out of range (0-{})", page_num, page_count - 1);
std::process::exit(1);
}
println!("Analyzing page {} of {}...\n", page_num + 1, page_count);
// Extract spans for detailed analysis
match doc.extract_spans(page_num) {
Ok(spans) => {
println!("=== EXTRACTED SPANS ===\n");
println!("Total spans: {}\n", spans.len());
// Analyze each span
for (i, span) in spans.iter().enumerate() {
let bold_status = if span.font_weight.is_bold() {
"BOLD"
} else {
"normal"
};
println!("[{}] Text: '{}'", i, span.text.chars().take(50).collect::<String>());
println!(" Font: {} {}pt {}", span.font_name, span.font_size, bold_status);
println!(
" Position: ({:.1}, {:.1}), Size: {:.1}×{:.1}",
span.bbox.x, span.bbox.y, span.bbox.width, span.bbox.height
);
println!(
" Color: RGB({:.2}, {:.2}, {:.2})",
span.color.r, span.color.g, span.color.b
);
println!();
}
// Analyze spacing between consecutive spans
println!("=== SPACING ANALYSIS ===\n");
println!("Analyzing gaps between consecutive spans on same line:\n");
let mut prev_idx: Option<usize> = None;
for (i, span) in spans.iter().enumerate() {
if let Some(prev_i) = prev_idx {
let prev = &spans[prev_i];
let y_diff = (prev.bbox.y - span.bbox.y).abs();
// Same-line tolerance scales with the larger of the two
// font sizes so superscripts and subscripts stay on the
// line they belong to (matches PdfDocument::same_line_threshold).
let line_tolerance = prev.font_size.max(span.font_size).max(1.0) * 0.5;
// Same-line check. Uses `<=` so the boundary case
// (y_diff exactly equal to the tolerance) is
// classified as same-line. Production treats a span
// as a new line only when `y_diff > threshold`, so
// the equality case belongs on the same-line side;
// this debug tool should report boundary deltas the
// same way.
if y_diff <= line_tolerance {
let gap = span.bbox.x - (prev.bbox.x + prev.bbox.width);
let space_threshold = prev.font_size * 0.25;
println!("Gap {} → {}: {:.2}pt", prev_i, i, gap);
println!(
" Spans: '{}' | '{}'",
prev.text.chars().take(30).collect::<String>(),
span.text.chars().take(30).collect::<String>()
);
println!(
" Fonts: {} {}pt | {} {}pt",
prev.font_name, prev.font_size, span.font_name, span.font_size
);
println!(
" Bold: {} | {}",
if prev.font_weight.is_bold() { "Y" } else { "N" },
if span.font_weight.is_bold() { "Y" } else { "N" }
);
println!(" Space threshold: {:.2}pt (font_size * 0.25)", space_threshold);
if gap < 0.0 {
println!(" ⚠️ NEGATIVE GAP: Spans overlap by {:.2}pt", -gap);
} else if gap < 0.1 {
println!(" ⚠️ Very small gap (< 0.1pt)");
} else if gap < space_threshold {
println!(
" ⚠️ Small gap ({:.2}pt < threshold {:.2}pt) - may merge without space",
gap, space_threshold
);
} else if gap < 3.0 {
println!(" ✓ Moderate gap ({:.2}pt) - should merge with space", gap);
} else {
println!(" ✓ Large gap ({:.2}pt) - separate spans", gap);
}
// Font transition detection
if prev.font_name != span.font_name {
println!(
" 🔴 Font transition: {} → {}",
prev.font_name, span.font_name
);
}
if (prev.font_size - span.font_size).abs() > 0.5 {
println!(
" 🔴 Size transition: {:.1}pt → {:.1}pt",
prev.font_size, span.font_size
);
}
if prev.font_weight.is_bold() != span.font_weight.is_bold() {
println!(
" 🔴 Bold transition: {} → {}",
if prev.font_weight.is_bold() {
"bold"
} else {
"normal"
},
if span.font_weight.is_bold() {
"bold"
} else {
"normal"
}
);
}
println!();
}
}
prev_idx = Some(i);
}
// Font analysis
println!("=== FONT ANALYSIS ===\n");
let mut fonts_used: std::collections::BTreeMap<String, usize> =
std::collections::BTreeMap::new();
let mut font_weights: std::collections::BTreeMap<String, Vec<String>> =
std::collections::BTreeMap::new();
for span in spans.iter() {
*fonts_used.entry(span.font_name.clone()).or_insert(0) += 1;
font_weights
.entry(span.font_name.clone())
.or_default()
.push(if span.font_weight.is_bold() {
"bold".to_string()
} else {
"normal".to_string()
});
}
for (font, count) in fonts_used {
let weights = &font_weights[&font];
let bold_count = weights.iter().filter(|w| w.as_str() == "bold").count();
println!("Font: {} (used {} times, {} bold spans)", font, count, bold_count);
}
println!();
},
Err(e) => {
eprintln!("Error extracting spans: {}", e);
std::process::exit(1);
},
}
Ok(())
}