rumdl/utils/
regex_cache.rs

1//!
2//! Cached Regex Patterns and Fast Content Checks for Markdown Linting
3//!
4//! This module provides a centralized collection of pre-compiled, cached regex patterns
5//! for all major Markdown constructs (headings, lists, code blocks, links, images, etc.).
6//! It also includes fast-path utility functions for quickly checking if content
7//! potentially contains certain Markdown elements, allowing rules to skip expensive
8//! processing when unnecessary.
9//!
10//! # Performance
11//!
12//! All regexes are compiled once at startup using `lazy_static`, avoiding repeated
13//! compilation and improving performance across the linter. Use these shared patterns
14//! in rules instead of compiling new regexes.
15//!
16//! # Usage
17//!
18//! - Use the provided statics for common Markdown patterns.
19//! - Use the `regex_lazy!` macro for ad-hoc regexes that are not predefined.
20//! - Use the utility functions for fast content checks before running regexes.
21
22use fancy_regex::Regex as FancyRegex;
23use lazy_static::lazy_static;
24use regex::Regex;
25use std::collections::HashMap;
26use std::sync::{Arc, Mutex};
27
28/// Global regex cache for dynamic patterns
29#[derive(Debug)]
30pub struct RegexCache {
31    cache: HashMap<String, Arc<Regex>>,
32    fancy_cache: HashMap<String, Arc<FancyRegex>>,
33    usage_stats: HashMap<String, u64>,
34}
35
36impl Default for RegexCache {
37    fn default() -> Self {
38        Self::new()
39    }
40}
41
42impl RegexCache {
43    pub fn new() -> Self {
44        Self {
45            cache: HashMap::new(),
46            fancy_cache: HashMap::new(),
47            usage_stats: HashMap::new(),
48        }
49    }
50
51    /// Get or compile a regex pattern
52    pub fn get_regex(&mut self, pattern: &str) -> Result<Arc<Regex>, regex::Error> {
53        if let Some(regex) = self.cache.get(pattern) {
54            *self.usage_stats.entry(pattern.to_string()).or_insert(0) += 1;
55            return Ok(regex.clone());
56        }
57
58        let regex = Arc::new(Regex::new(pattern)?);
59        self.cache.insert(pattern.to_string(), regex.clone());
60        *self.usage_stats.entry(pattern.to_string()).or_insert(0) += 1;
61        Ok(regex)
62    }
63
64    /// Get or compile a fancy regex pattern
65    pub fn get_fancy_regex(&mut self, pattern: &str) -> Result<Arc<FancyRegex>, Box<fancy_regex::Error>> {
66        if let Some(regex) = self.fancy_cache.get(pattern) {
67            *self.usage_stats.entry(pattern.to_string()).or_insert(0) += 1;
68            return Ok(regex.clone());
69        }
70
71        match FancyRegex::new(pattern) {
72            Ok(regex) => {
73                let arc_regex = Arc::new(regex);
74                self.fancy_cache.insert(pattern.to_string(), arc_regex.clone());
75                *self.usage_stats.entry(pattern.to_string()).or_insert(0) += 1;
76                Ok(arc_regex)
77            }
78            Err(e) => Err(Box::new(e)),
79        }
80    }
81
82    /// Get cache statistics
83    pub fn get_stats(&self) -> HashMap<String, u64> {
84        self.usage_stats.clone()
85    }
86
87    /// Clear cache (useful for testing)
88    pub fn clear(&mut self) {
89        self.cache.clear();
90        self.fancy_cache.clear();
91        self.usage_stats.clear();
92    }
93}
94
95lazy_static! {
96    /// Global regex cache instance
97    static ref GLOBAL_REGEX_CACHE: Arc<Mutex<RegexCache>> = Arc::new(Mutex::new(RegexCache::new()));
98}
99
100/// Get a regex from the global cache
101pub fn get_cached_regex(pattern: &str) -> Result<Arc<Regex>, regex::Error> {
102    let mut cache = GLOBAL_REGEX_CACHE.lock().unwrap();
103    cache.get_regex(pattern)
104}
105
106/// Get a fancy regex from the global cache
107pub fn get_cached_fancy_regex(pattern: &str) -> Result<Arc<FancyRegex>, Box<fancy_regex::Error>> {
108    let mut cache = GLOBAL_REGEX_CACHE.lock().unwrap();
109    cache.get_fancy_regex(pattern)
110}
111
112/// Get cache usage statistics
113pub fn get_cache_stats() -> HashMap<String, u64> {
114    let cache = GLOBAL_REGEX_CACHE.lock().unwrap();
115    cache.get_stats()
116}
117
118/// Macro for defining a lazily-initialized, cached regex pattern.
119/// Use this for ad-hoc regexes that are not already defined in this module.
120/// Example:
121/// ```
122/// use rumdl::regex_lazy;
123/// let my_re = regex_lazy!(r"^foo.*bar$");
124/// assert!(my_re.is_match("foobar"));
125/// ```
126#[macro_export]
127macro_rules! regex_lazy {
128    ($pattern:expr) => {{
129        lazy_static::lazy_static! {
130            static ref REGEX: regex::Regex = regex::Regex::new($pattern).unwrap();
131        }
132        &*REGEX
133    }};
134}
135
136/// Macro for getting regex from global cache
137#[macro_export]
138macro_rules! regex_cached {
139    ($pattern:expr) => {{ $crate::utils::regex_cache::get_cached_regex($pattern).expect("Failed to compile regex") }};
140}
141
142/// Macro for getting fancy regex from global cache
143#[macro_export]
144macro_rules! fancy_regex_cached {
145    ($pattern:expr) => {{ $crate::utils::regex_cache::get_cached_fancy_regex($pattern).expect("Failed to compile fancy regex") }};
146}
147
148// Also make the macro available directly from this module
149pub use crate::regex_lazy;
150
151lazy_static! {
152    // URL patterns
153    pub static ref URL_REGEX: Regex = Regex::new(r#"(?:https?|ftp)://[^\s<>\[\]()'"]+[^\s<>\[\]()"'.,]"#).unwrap();
154    pub static ref BARE_URL_REGEX: Regex = Regex::new(r"(?:https?|ftp)://[^\s<>]+[^\s<>.]").unwrap();
155    pub static ref URL_PATTERN: Regex = Regex::new(r"((?:https?|ftp)://[^\s\)<>]+[^\s\)<>.,])").unwrap();
156
157    // Heading patterns
158    pub static ref ATX_HEADING_REGEX: Regex = Regex::new(r"^(\s*)(#{1,6})(\s+|$)").unwrap();
159    pub static ref CLOSED_ATX_HEADING_REGEX: Regex = Regex::new(r"^(\s*)(#{1,6})(\s+)(.*)(\s+)(#+)(\s*)$").unwrap();
160    pub static ref SETEXT_HEADING_REGEX: Regex = Regex::new(r"^(\s*)[^\s]+.*\n(\s*)(=+|-+)\s*$").unwrap();
161    pub static ref TRAILING_PUNCTUATION_REGEX: Regex = Regex::new(r"[.,:;!?]$").unwrap();
162
163    // ATX heading patterns for MD051 and other rules
164    pub static ref ATX_HEADING_WITH_CAPTURE: Regex = Regex::new(r"^(#{1,6})\s+(.+?)(?:\s+#*\s*)?$").unwrap();
165    pub static ref SETEXT_HEADING_WITH_CAPTURE: FancyRegex = FancyRegex::new(r"^([^\n]+)\n([=\-])\2+\s*$").unwrap();
166
167    // List patterns
168    pub static ref UNORDERED_LIST_MARKER_REGEX: Regex = Regex::new(r"^(\s*)([*+-])(\s+)").unwrap();
169    pub static ref ORDERED_LIST_MARKER_REGEX: Regex = Regex::new(r"^(\s*)(\d+)([.)])(\s+)").unwrap();
170    pub static ref LIST_MARKER_ANY_REGEX: Regex = Regex::new(r"^(\s*)(?:([*+-])|(\d+)[.)])(\s+)").unwrap();
171
172    // Code block patterns
173    pub static ref FENCED_CODE_BLOCK_START_REGEX: Regex = Regex::new(r"^(\s*)(```|~~~)(.*)$").unwrap();
174    pub static ref FENCED_CODE_BLOCK_END_REGEX: Regex = Regex::new(r"^(\s*)(```|~~~)(\s*)$").unwrap();
175    pub static ref INDENTED_CODE_BLOCK_REGEX: Regex = Regex::new(r"^(\s{4,})(.*)$").unwrap();
176    pub static ref CODE_FENCE_REGEX: Regex = Regex::new(r"^(`{3,}|~{3,})").unwrap();
177
178    // Emphasis patterns
179    pub static ref EMPHASIS_REGEX: FancyRegex = FancyRegex::new(r"(\s|^)(\*{1,2}|_{1,2})(?=\S)(.+?)(?<=\S)(\2)(\s|$)").unwrap();
180    pub static ref SPACE_IN_EMPHASIS_REGEX: FancyRegex = FancyRegex::new(r"(\*|_)(\s+)(.+?)(\s+)(\1)").unwrap();
181
182    // MD037 specific emphasis patterns - improved to avoid false positives
183    // Only match emphasis with spaces that are actually complete emphasis blocks
184    // Use word boundaries and negative lookbehind/lookahead to avoid matching across emphasis boundaries
185    pub static ref ASTERISK_EMPHASIS: Regex = Regex::new(r"(?:^|[^*])\*(\s+[^*]+\s*|\s*[^*]+\s+)\*(?:[^*]|$)").unwrap();
186    pub static ref UNDERSCORE_EMPHASIS: Regex = Regex::new(r"(?:^|[^_])_(\s+[^_]+\s*|\s*[^_]+\s+)_(?:[^_]|$)").unwrap();
187    pub static ref DOUBLE_UNDERSCORE_EMPHASIS: Regex = Regex::new(r"(?:^|[^_])__(\s+[^_]+\s*|\s*[^_]+\s+)__(?:[^_]|$)").unwrap();
188    pub static ref DOUBLE_ASTERISK_EMPHASIS: FancyRegex = FancyRegex::new(r"\*\*\s+([^*]+?)\s+\*\*").unwrap();
189    pub static ref DOUBLE_ASTERISK_SPACE_START: FancyRegex = FancyRegex::new(r"\*\*\s+([^*]+?)\*\*").unwrap();
190    pub static ref DOUBLE_ASTERISK_SPACE_END: FancyRegex = FancyRegex::new(r"\*\*([^*]+?)\s+\*\*").unwrap();
191
192    // Code block patterns
193    pub static ref FENCED_CODE_BLOCK_START: Regex = Regex::new(r"^(\s*)```(?:[^`\r\n]*)$").unwrap();
194    pub static ref FENCED_CODE_BLOCK_END: Regex = Regex::new(r"^(\s*)```\s*$").unwrap();
195    pub static ref ALTERNATE_FENCED_CODE_BLOCK_START: Regex = Regex::new(r"^(\s*)~~~(?:[^~\r\n]*)$").unwrap();
196    pub static ref ALTERNATE_FENCED_CODE_BLOCK_END: Regex = Regex::new(r"^(\s*)~~~\s*$").unwrap();
197    pub static ref INDENTED_CODE_BLOCK_PATTERN: Regex = Regex::new(r"^(\s{4,})").unwrap();
198
199    // HTML patterns
200    pub static ref HTML_TAG_REGEX: Regex = Regex::new(r"<([a-zA-Z][^>]*)>").unwrap();
201    pub static ref HTML_SELF_CLOSING_TAG_REGEX: Regex = Regex::new(r"<([a-zA-Z][^>]*/)>").unwrap();
202    pub static ref HTML_TAG_FINDER: Regex = Regex::new("(?i)</?[a-zA-Z][^>]*>").unwrap();
203    pub static ref HTML_TAG_QUICK_CHECK: Regex = Regex::new("(?i)</?[a-zA-Z]").unwrap();
204
205    // Link patterns for MD051 and other rules
206    pub static ref LINK_REFERENCE_DEFINITION_REGEX: Regex = Regex::new(r"^\s*\[([^\]]+)\]:\s+(.+)$").unwrap();
207    pub static ref INLINE_LINK_REGEX: Regex = Regex::new(r"\[([^\]]+)\]\(([^)]+)\)").unwrap();
208    pub static ref LINK_TEXT_REGEX: Regex = Regex::new(r"\[([^\]]*)\]").unwrap();
209    pub static ref LINK_REGEX: FancyRegex = FancyRegex::new(r"(?<!\\)\[([^\]]*)\]\(([^)#]*)#([^)]+)\)").unwrap();
210    pub static ref EXTERNAL_URL_REGEX: FancyRegex = FancyRegex::new(r"^(https?://|ftp://|www\.|[^/]+\.[a-z]{2,})").unwrap();
211
212    // Image patterns
213    pub static ref IMAGE_REGEX: Regex = Regex::new(r"!\[([^\]]*)\]\(([^)]+)\)").unwrap();
214
215    // Whitespace patterns
216    pub static ref TRAILING_WHITESPACE_REGEX: Regex = Regex::new(r"\s+$").unwrap();
217    pub static ref MULTIPLE_BLANK_LINES_REGEX: Regex = Regex::new(r"\n{3,}").unwrap();
218
219    // Front matter patterns
220    pub static ref FRONT_MATTER_REGEX: Regex = Regex::new(r"^---\n.*?\n---\n").unwrap();
221
222    // MD051 specific patterns
223    pub static ref INLINE_CODE_REGEX: FancyRegex = FancyRegex::new(r"`[^`]+`").unwrap();
224    pub static ref BOLD_ASTERISK_REGEX: Regex = Regex::new(r"\*\*(.+?)\*\*").unwrap();
225    pub static ref BOLD_UNDERSCORE_REGEX: Regex = Regex::new(r"__(.+?)__").unwrap();
226    pub static ref ITALIC_ASTERISK_REGEX: Regex = Regex::new(r"\*([^*]+?)\*").unwrap();
227    pub static ref ITALIC_UNDERSCORE_REGEX: Regex = Regex::new(r"_([^_]+?)_").unwrap();
228    pub static ref LINK_TEXT_FULL_REGEX: FancyRegex = FancyRegex::new(r"\[([^\]]*)\]\([^)]*\)").unwrap();
229    pub static ref STRIKETHROUGH_REGEX: Regex = Regex::new(r"~~(.+?)~~").unwrap();
230    pub static ref MULTIPLE_HYPHENS: Regex = Regex::new(r"-{2,}").unwrap();
231    pub static ref TOC_SECTION_START: Regex = Regex::new(r"^#+\s*(?:Table of Contents|Contents|TOC)\s*$").unwrap();
232
233    // Blockquote patterns
234    pub static ref BLOCKQUOTE_PREFIX_RE: Regex = Regex::new(r"^(\s*>+\s*)").unwrap();
235}
236
237/// Utility functions for quick content checks
238/// Check if content contains any headings (quick check before regex)
239pub fn has_heading_markers(content: &str) -> bool {
240    content.contains('#')
241}
242
243/// Check if content contains any lists (quick check before regex)
244pub fn has_list_markers(content: &str) -> bool {
245    content.contains('*')
246        || content.contains('-')
247        || content.contains('+')
248        || (content.contains('.') && content.contains(|c: char| c.is_ascii_digit()))
249}
250
251/// Check if content contains any code blocks (quick check before regex)
252pub fn has_code_block_markers(content: &str) -> bool {
253    content.contains("```") || content.contains("~~~") || content.contains("\n    ")
254    // Indented code block potential
255}
256
257/// Check if content contains any emphasis markers (quick check before regex)
258pub fn has_emphasis_markers(content: &str) -> bool {
259    content.contains('*') || content.contains('_')
260}
261
262/// Check if content contains any HTML tags (quick check before regex)
263pub fn has_html_tags(content: &str) -> bool {
264    content.contains('<') && (content.contains('>') || content.contains("/>"))
265}
266
267/// Check if content contains any links (quick check before regex)
268pub fn has_link_markers(content: &str) -> bool {
269    (content.contains('[') && content.contains(']'))
270        || content.contains("http://")
271        || content.contains("https://")
272        || content.contains("ftp://")
273}
274
275/// Check if content contains any images (quick check before regex)
276pub fn has_image_markers(content: &str) -> bool {
277    content.contains("![")
278}
279
280/// Optimize URL detection by implementing a character-by-character scanner
281/// that's much faster than regex for cases where we know there's no URL
282pub fn contains_url(content: &str) -> bool {
283    // Fast check - if these substrings aren't present, there's no URL
284    if !content.contains("://") {
285        return false;
286    }
287
288    let chars: Vec<char> = content.chars().collect();
289    let mut i = 0;
290
291    while i < chars.len() {
292        // Look for the start of a URL protocol
293        if i + 2 < chars.len()
294            && ((chars[i] == 'h' && chars[i + 1] == 't' && chars[i + 2] == 't')
295                || (chars[i] == 'f' && chars[i + 1] == 't' && chars[i + 2] == 'p'))
296        {
297            // Scan forward to find "://"
298            let mut j = i;
299            while j + 2 < chars.len() {
300                if chars[j] == ':' && chars[j + 1] == '/' && chars[j + 2] == '/' {
301                    return true;
302                }
303                j += 1;
304
305                // Don't scan too far ahead for the protocol
306                if j > i + 10 {
307                    break;
308                }
309            }
310        }
311        i += 1;
312    }
313
314    false
315}
316
317/// Escapes a string to be used in a regex pattern
318pub fn escape_regex(s: &str) -> String {
319    let special_chars = ['.', '+', '*', '?', '^', '$', '(', ')', '[', ']', '{', '}', '|', '\\'];
320    let mut result = String::with_capacity(s.len() * 2);
321
322    for c in s.chars() {
323        if special_chars.contains(&c) {
324            result.push('\\');
325        }
326        result.push(c);
327    }
328
329    result
330}
331
332#[cfg(test)]
333mod tests {
334    use super::*;
335
336    #[test]
337    fn test_regex_cache_new() {
338        let cache = RegexCache::new();
339        assert!(cache.cache.is_empty());
340        assert!(cache.fancy_cache.is_empty());
341        assert!(cache.usage_stats.is_empty());
342    }
343
344    #[test]
345    fn test_regex_cache_default() {
346        let cache = RegexCache::default();
347        assert!(cache.cache.is_empty());
348        assert!(cache.fancy_cache.is_empty());
349        assert!(cache.usage_stats.is_empty());
350    }
351
352    #[test]
353    fn test_get_regex_compilation() {
354        let mut cache = RegexCache::new();
355
356        // First call compiles and caches
357        let regex1 = cache.get_regex(r"\d+").unwrap();
358        assert_eq!(cache.cache.len(), 1);
359        assert_eq!(cache.usage_stats.get(r"\d+"), Some(&1));
360
361        // Second call returns cached version
362        let regex2 = cache.get_regex(r"\d+").unwrap();
363        assert_eq!(cache.cache.len(), 1);
364        assert_eq!(cache.usage_stats.get(r"\d+"), Some(&2));
365
366        // Both should be the same Arc
367        assert!(Arc::ptr_eq(&regex1, &regex2));
368    }
369
370    #[test]
371    fn test_get_regex_invalid_pattern() {
372        let mut cache = RegexCache::new();
373        let result = cache.get_regex(r"[unterminated");
374        assert!(result.is_err());
375        assert!(cache.cache.is_empty());
376    }
377
378    #[test]
379    fn test_get_fancy_regex_compilation() {
380        let mut cache = RegexCache::new();
381
382        // First call compiles and caches
383        let regex1 = cache.get_fancy_regex(r"(?<=foo)bar").unwrap();
384        assert_eq!(cache.fancy_cache.len(), 1);
385        assert_eq!(cache.usage_stats.get(r"(?<=foo)bar"), Some(&1));
386
387        // Second call returns cached version
388        let regex2 = cache.get_fancy_regex(r"(?<=foo)bar").unwrap();
389        assert_eq!(cache.fancy_cache.len(), 1);
390        assert_eq!(cache.usage_stats.get(r"(?<=foo)bar"), Some(&2));
391
392        // Both should be the same Arc
393        assert!(Arc::ptr_eq(&regex1, &regex2));
394    }
395
396    #[test]
397    fn test_get_fancy_regex_invalid_pattern() {
398        let mut cache = RegexCache::new();
399        let result = cache.get_fancy_regex(r"(?<=invalid");
400        assert!(result.is_err());
401        assert!(cache.fancy_cache.is_empty());
402    }
403
404    #[test]
405    fn test_get_stats() {
406        let mut cache = RegexCache::new();
407
408        // Use some patterns
409        let _ = cache.get_regex(r"\d+").unwrap();
410        let _ = cache.get_regex(r"\d+").unwrap();
411        let _ = cache.get_regex(r"\w+").unwrap();
412        let _ = cache.get_fancy_regex(r"(?<=foo)bar").unwrap();
413
414        let stats = cache.get_stats();
415        assert_eq!(stats.get(r"\d+"), Some(&2));
416        assert_eq!(stats.get(r"\w+"), Some(&1));
417        assert_eq!(stats.get(r"(?<=foo)bar"), Some(&1));
418    }
419
420    #[test]
421    fn test_clear_cache() {
422        let mut cache = RegexCache::new();
423
424        // Add some patterns
425        let _ = cache.get_regex(r"\d+").unwrap();
426        let _ = cache.get_fancy_regex(r"(?<=foo)bar").unwrap();
427
428        assert!(!cache.cache.is_empty());
429        assert!(!cache.fancy_cache.is_empty());
430        assert!(!cache.usage_stats.is_empty());
431
432        // Clear cache
433        cache.clear();
434
435        assert!(cache.cache.is_empty());
436        assert!(cache.fancy_cache.is_empty());
437        assert!(cache.usage_stats.is_empty());
438    }
439
440    #[test]
441    fn test_global_cache_functions() {
442        // Test get_cached_regex
443        let regex1 = get_cached_regex(r"\d{3}").unwrap();
444        let regex2 = get_cached_regex(r"\d{3}").unwrap();
445        assert!(Arc::ptr_eq(&regex1, &regex2));
446
447        // Test get_cached_fancy_regex
448        let fancy1 = get_cached_fancy_regex(r"(?<=test)ing").unwrap();
449        let fancy2 = get_cached_fancy_regex(r"(?<=test)ing").unwrap();
450        assert!(Arc::ptr_eq(&fancy1, &fancy2));
451
452        // Test stats
453        let stats = get_cache_stats();
454        assert!(stats.contains_key(r"\d{3}"));
455        assert!(stats.contains_key(r"(?<=test)ing"));
456    }
457
458    #[test]
459    fn test_regex_lazy_macro() {
460        let re = regex_lazy!(r"^test.*end$");
461        assert!(re.is_match("test something end"));
462        assert!(!re.is_match("test something"));
463
464        // The macro creates a new static for each invocation location,
465        // so we can't test pointer equality across different invocations
466        // But we can test that the regex works correctly
467        let re2 = regex_lazy!(r"^start.*finish$");
468        assert!(re2.is_match("start and finish"));
469        assert!(!re2.is_match("start without end"));
470    }
471
472    #[test]
473    fn test_has_heading_markers() {
474        assert!(has_heading_markers("# Heading"));
475        assert!(has_heading_markers("Text with # symbol"));
476        assert!(!has_heading_markers("Text without heading marker"));
477    }
478
479    #[test]
480    fn test_has_list_markers() {
481        assert!(has_list_markers("* Item"));
482        assert!(has_list_markers("- Item"));
483        assert!(has_list_markers("+ Item"));
484        assert!(has_list_markers("1. Item"));
485        assert!(!has_list_markers("Text without list markers"));
486    }
487
488    #[test]
489    fn test_has_code_block_markers() {
490        assert!(has_code_block_markers("```code```"));
491        assert!(has_code_block_markers("~~~code~~~"));
492        assert!(has_code_block_markers("Text\n    indented code"));
493        assert!(!has_code_block_markers("Text without code blocks"));
494    }
495
496    #[test]
497    fn test_has_emphasis_markers() {
498        assert!(has_emphasis_markers("*emphasis*"));
499        assert!(has_emphasis_markers("_emphasis_"));
500        assert!(has_emphasis_markers("**bold**"));
501        assert!(has_emphasis_markers("__bold__"));
502        assert!(!has_emphasis_markers("no emphasis"));
503    }
504
505    #[test]
506    fn test_has_html_tags() {
507        assert!(has_html_tags("<div>content</div>"));
508        assert!(has_html_tags("<br/>"));
509        assert!(has_html_tags("<img src='test.jpg'>"));
510        assert!(!has_html_tags("no html tags"));
511        assert!(!has_html_tags("less than < but no tag"));
512    }
513
514    #[test]
515    fn test_has_link_markers() {
516        assert!(has_link_markers("[text](url)"));
517        assert!(has_link_markers("[reference][1]"));
518        assert!(has_link_markers("http://example.com"));
519        assert!(has_link_markers("https://example.com"));
520        assert!(has_link_markers("ftp://example.com"));
521        assert!(!has_link_markers("no links here"));
522    }
523
524    #[test]
525    fn test_has_image_markers() {
526        assert!(has_image_markers("![alt text](image.png)"));
527        assert!(has_image_markers("![](image.png)"));
528        assert!(!has_image_markers("[link](url)"));
529        assert!(!has_image_markers("no images"));
530    }
531
532    #[test]
533    fn test_contains_url() {
534        assert!(contains_url("http://example.com"));
535        assert!(contains_url("Text with https://example.com link"));
536        assert!(contains_url("ftp://example.com"));
537        assert!(!contains_url("Text without URL"));
538        assert!(!contains_url("http not followed by ://"));
539
540        // Edge cases
541        assert!(!contains_url("http"));
542        assert!(!contains_url("https"));
543        assert!(!contains_url("://"));
544        assert!(contains_url("Visit http://site.com now"));
545        assert!(contains_url("See https://secure.site.com/path"));
546    }
547
548    #[test]
549    fn test_contains_url_performance() {
550        // Test early exit for strings without "://"
551        let long_text = "a".repeat(10000);
552        assert!(!contains_url(&long_text));
553
554        // Test with URL at the end
555        let text_with_url = format!("{long_text}https://example.com");
556        assert!(contains_url(&text_with_url));
557    }
558
559    #[test]
560    fn test_escape_regex() {
561        assert_eq!(escape_regex("a.b"), "a\\.b");
562        assert_eq!(escape_regex("a+b*c"), "a\\+b\\*c");
563        assert_eq!(escape_regex("(test)"), "\\(test\\)");
564        assert_eq!(escape_regex("[a-z]"), "\\[a-z\\]");
565        assert_eq!(escape_regex("normal text"), "normal text");
566
567        // Test all special characters
568        assert_eq!(escape_regex(".$^{[(|)*+?\\"), "\\.\\$\\^\\{\\[\\(\\|\\)\\*\\+\\?\\\\");
569
570        // Test empty string
571        assert_eq!(escape_regex(""), "");
572
573        // Test mixed content
574        assert_eq!(escape_regex("test.com/path?query=1"), "test\\.com/path\\?query=1");
575    }
576
577    #[test]
578    fn test_static_regex_patterns() {
579        // Test URL patterns
580        assert!(URL_REGEX.is_match("https://example.com"));
581        assert!(URL_REGEX.is_match("http://test.org/path"));
582        assert!(URL_REGEX.is_match("ftp://files.com"));
583        assert!(!URL_REGEX.is_match("not a url"));
584
585        // Test heading patterns
586        assert!(ATX_HEADING_REGEX.is_match("# Heading"));
587        assert!(ATX_HEADING_REGEX.is_match("  ## Indented"));
588        assert!(ATX_HEADING_REGEX.is_match("### "));
589        assert!(!ATX_HEADING_REGEX.is_match("Not a heading"));
590
591        // Test list patterns
592        assert!(UNORDERED_LIST_MARKER_REGEX.is_match("* Item"));
593        assert!(UNORDERED_LIST_MARKER_REGEX.is_match("- Item"));
594        assert!(UNORDERED_LIST_MARKER_REGEX.is_match("+ Item"));
595        assert!(ORDERED_LIST_MARKER_REGEX.is_match("1. Item"));
596        assert!(ORDERED_LIST_MARKER_REGEX.is_match("99. Item"));
597
598        // Test code block patterns
599        assert!(FENCED_CODE_BLOCK_START_REGEX.is_match("```"));
600        assert!(FENCED_CODE_BLOCK_START_REGEX.is_match("```rust"));
601        assert!(FENCED_CODE_BLOCK_START_REGEX.is_match("~~~"));
602        assert!(FENCED_CODE_BLOCK_END_REGEX.is_match("```"));
603        assert!(FENCED_CODE_BLOCK_END_REGEX.is_match("~~~"));
604
605        // Test emphasis patterns
606        assert!(BOLD_ASTERISK_REGEX.is_match("**bold**"));
607        assert!(BOLD_UNDERSCORE_REGEX.is_match("__bold__"));
608        assert!(ITALIC_ASTERISK_REGEX.is_match("*italic*"));
609        assert!(ITALIC_UNDERSCORE_REGEX.is_match("_italic_"));
610
611        // Test HTML patterns
612        assert!(HTML_TAG_REGEX.is_match("<div>"));
613        assert!(HTML_TAG_REGEX.is_match("<span class='test'>"));
614        assert!(HTML_SELF_CLOSING_TAG_REGEX.is_match("<br/>"));
615        assert!(HTML_SELF_CLOSING_TAG_REGEX.is_match("<img src='test'/>"));
616
617        // Test whitespace patterns
618        assert!(TRAILING_WHITESPACE_REGEX.is_match("line with spaces   "));
619        assert!(TRAILING_WHITESPACE_REGEX.is_match("tabs\t\t"));
620        assert!(MULTIPLE_BLANK_LINES_REGEX.is_match("\n\n\n"));
621        assert!(MULTIPLE_BLANK_LINES_REGEX.is_match("\n\n\n\n"));
622
623        // Test blockquote pattern
624        assert!(BLOCKQUOTE_PREFIX_RE.is_match("> Quote"));
625        assert!(BLOCKQUOTE_PREFIX_RE.is_match("  > Indented quote"));
626        assert!(BLOCKQUOTE_PREFIX_RE.is_match(">> Nested"));
627    }
628
629    #[test]
630    fn test_thread_safety() {
631        use std::thread;
632
633        let handles: Vec<_> = (0..10)
634            .map(|i| {
635                thread::spawn(move || {
636                    let pattern = format!(r"\d{{{i}}}");
637                    let regex = get_cached_regex(&pattern).unwrap();
638                    assert!(regex.is_match(&"1".repeat(i)));
639                })
640            })
641            .collect();
642
643        for handle in handles {
644            handle.join().unwrap();
645        }
646    }
647}
rumdl/utils/regex_cache.rs

rumdl/utils/
regex_cache.rs
