@agentsean/crawler 1.0.0

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (70) hide show
  1. package/LICENSE +661 -0
  2. package/dist/adaptive.d.ts +9 -0
  3. package/dist/adaptive.d.ts.map +1 -0
  4. package/dist/adaptive.js +43 -0
  5. package/dist/adaptive.js.map +1 -0
  6. package/dist/crawl.d.ts +4 -0
  7. package/dist/crawl.d.ts.map +1 -0
  8. package/dist/crawl.js +348 -0
  9. package/dist/crawl.js.map +1 -0
  10. package/dist/decode.d.ts +7 -0
  11. package/dist/decode.d.ts.map +1 -0
  12. package/dist/decode.js +53 -0
  13. package/dist/decode.js.map +1 -0
  14. package/dist/extract.d.ts +4 -0
  15. package/dist/extract.d.ts.map +1 -0
  16. package/dist/extract.js +268 -0
  17. package/dist/extract.js.map +1 -0
  18. package/dist/hash.d.ts +4 -0
  19. package/dist/hash.d.ts.map +1 -0
  20. package/dist/hash.js +11 -0
  21. package/dist/hash.js.map +1 -0
  22. package/dist/http.d.ts +38 -0
  23. package/dist/http.d.ts.map +1 -0
  24. package/dist/http.js +207 -0
  25. package/dist/http.js.map +1 -0
  26. package/dist/index.d.ts +14 -0
  27. package/dist/index.d.ts.map +1 -0
  28. package/dist/index.js +13 -0
  29. package/dist/index.js.map +1 -0
  30. package/dist/persist.d.ts +20 -0
  31. package/dist/persist.d.ts.map +1 -0
  32. package/dist/persist.js +186 -0
  33. package/dist/persist.js.map +1 -0
  34. package/dist/probe.d.ts +4 -0
  35. package/dist/probe.d.ts.map +1 -0
  36. package/dist/probe.js +132 -0
  37. package/dist/probe.js.map +1 -0
  38. package/dist/readability.d.ts +9 -0
  39. package/dist/readability.d.ts.map +1 -0
  40. package/dist/readability.js +60 -0
  41. package/dist/readability.js.map +1 -0
  42. package/dist/render.d.ts +15 -0
  43. package/dist/render.d.ts.map +1 -0
  44. package/dist/render.js +87 -0
  45. package/dist/render.js.map +1 -0
  46. package/dist/robots.d.ts +13 -0
  47. package/dist/robots.d.ts.map +1 -0
  48. package/dist/robots.js +187 -0
  49. package/dist/robots.js.map +1 -0
  50. package/dist/simhash.d.ts +9 -0
  51. package/dist/simhash.d.ts.map +1 -0
  52. package/dist/simhash.js +58 -0
  53. package/dist/simhash.js.map +1 -0
  54. package/dist/sitemap.d.ts +3 -0
  55. package/dist/sitemap.d.ts.map +1 -0
  56. package/dist/sitemap.js +165 -0
  57. package/dist/sitemap.js.map +1 -0
  58. package/dist/types.d.ts +251 -0
  59. package/dist/types.d.ts.map +1 -0
  60. package/dist/types.js +2 -0
  61. package/dist/types.js.map +1 -0
  62. package/dist/ua.d.ts +3 -0
  63. package/dist/ua.d.ts.map +1 -0
  64. package/dist/ua.js +3 -0
  65. package/dist/ua.js.map +1 -0
  66. package/dist/url.d.ts +11 -0
  67. package/dist/url.d.ts.map +1 -0
  68. package/dist/url.js +114 -0
  69. package/dist/url.js.map +1 -0
  70. package/package.json +60 -0
@@ -0,0 +1,251 @@
1
+ export type RedirectHop = {
2
+ url: string;
3
+ status: number;
4
+ location: string | null;
5
+ };
6
+ export type ExtractedLink = {
7
+ href: string;
8
+ absUrl: string;
9
+ anchorText: string;
10
+ rel: string[];
11
+ target: string | null;
12
+ isNofollow: boolean;
13
+ isUgc: boolean;
14
+ isSponsored: boolean;
15
+ isInternal: boolean;
16
+ position: "nav" | "header" | "footer" | "aside" | "main" | "unknown";
17
+ };
18
+ export type ExtractedImage = {
19
+ src: string;
20
+ absUrl: string;
21
+ alt: string | null;
22
+ hasAltAttr: boolean;
23
+ width: string | null;
24
+ height: string | null;
25
+ loading: string | null;
26
+ srcset: string | null;
27
+ decoding: string | null;
28
+ };
29
+ export type HreflangAlt = {
30
+ lang: string;
31
+ href: string;
32
+ absUrl: string;
33
+ inHead: boolean;
34
+ source: "head" | "header" | "body";
35
+ };
36
+ export type JsonLdBlock = {
37
+ raw: string;
38
+ parsed: unknown;
39
+ error: string | null;
40
+ inHead: boolean;
41
+ };
42
+ export type PageExtract = {
43
+ title: string | null;
44
+ titles: string[];
45
+ titleInHead: boolean;
46
+ metaDescription: string | null;
47
+ metaDescriptions: string[];
48
+ metaDescInHead: boolean;
49
+ h1: string[];
50
+ h2: string[];
51
+ headings: {
52
+ level: number;
53
+ text: string;
54
+ }[];
55
+ canonicalHtml: string[];
56
+ canonicalHeader: string[];
57
+ canonicalInHead: boolean;
58
+ canonicalOutsideHead: boolean;
59
+ headBrokenEarly: boolean;
60
+ robotsMeta: string[];
61
+ robotsMetaInHead: boolean;
62
+ xRobotsTag: string | null;
63
+ viewport: string | null;
64
+ lang: string | null;
65
+ charset: string | null;
66
+ charsetInFirst1024: boolean;
67
+ hreflang: HreflangAlt[];
68
+ jsonLd: JsonLdBlock[];
69
+ openGraph: Record<string, string>;
70
+ twitterCard: Record<string, string>;
71
+ links: ExtractedLink[];
72
+ images: ExtractedImage[];
73
+ scripts: {
74
+ src: string | null;
75
+ absUrl: string | null;
76
+ async: boolean;
77
+ defer: boolean;
78
+ }[];
79
+ stylesheets: {
80
+ href: string | null;
81
+ absUrl: string | null;
82
+ }[];
83
+ wordCount: number;
84
+ mainWordCount: number;
85
+ mainText: string;
86
+ allText: string;
87
+ boilerplateRatio: number;
88
+ metaRefresh: string | null;
89
+ hasNoscriptJsWarning: boolean;
90
+ spaRootEmpty: boolean;
91
+ hashFragmentMeta: boolean;
92
+ forms: {
93
+ action: string;
94
+ absAction: string;
95
+ }[];
96
+ mixedActive: string[];
97
+ mixedPassive: string[];
98
+ unsafeTargetBlanks: number;
99
+ plugins: boolean;
100
+ loremIpsum: boolean;
101
+ paginationRel: {
102
+ next: string | null;
103
+ prev: string | null;
104
+ };
105
+ htmlLang: string | null;
106
+ };
107
+ export type CrawledPage = {
108
+ url: string;
109
+ finalUrl: string;
110
+ statusCode: number | null;
111
+ error: string | null;
112
+ contentType: string;
113
+ headers: Record<string, string>;
114
+ ttfbMs: number;
115
+ totalMs: number;
116
+ wireBytes: number;
117
+ decodedBytes: number;
118
+ httpVersion: string | null;
119
+ redirectChain: RedirectHop[];
120
+ redirectLoop: boolean;
121
+ exceedsGoogleRedirectLimit: boolean;
122
+ html: string | null;
123
+ renderedHtml: string | null;
124
+ extract: PageExtract | null;
125
+ renderedExtract: PageExtract | null;
126
+ contentHash: string | null;
127
+ simhash: string | null;
128
+ etag: string | null;
129
+ lastModified: string | null;
130
+ notModified: boolean;
131
+ supports304: boolean | null;
132
+ inSitemap: boolean;
133
+ robotsAllowed: boolean;
134
+ depth: number;
135
+ inlinkCount: number;
136
+ outlinkCount: number;
137
+ followInlinkCount: number;
138
+ isInternal: boolean;
139
+ fetchedAt: string;
140
+ renderPolicy: "always" | "never" | "sample" | "skipped";
141
+ rendered: boolean;
142
+ jsDependencyScore: number;
143
+ templateKey: string;
144
+ };
145
+ export type RobotsOutcome = {
146
+ statusCode: number | null;
147
+ raw: string;
148
+ contentType: string;
149
+ bytes: number;
150
+ redirectHops: number;
151
+ error: string | null;
152
+ groups: RobotsGroup[];
153
+ sitemaps: string[];
154
+ unknownDirectives: string[];
155
+ hasNoindexDirective: boolean;
156
+ hasCrawlDelay: boolean;
157
+ crawlDelay: number | null;
158
+ disallowAll: boolean;
159
+ hash: string;
160
+ mode: "allow-all" | "disallow-all" | "parsed";
161
+ };
162
+ export type RobotsGroup = {
163
+ agents: string[];
164
+ rules: {
165
+ type: "allow" | "disallow";
166
+ pattern: string;
167
+ }[];
168
+ crawlDelay: number | null;
169
+ };
170
+ export type SitemapUrl = {
171
+ loc: string;
172
+ lastmod: string | null;
173
+ changefreq: string | null;
174
+ priority: string | null;
175
+ images: string[];
176
+ videoValid: boolean | null;
177
+ hreflang: {
178
+ lang: string;
179
+ href: string;
180
+ }[];
181
+ };
182
+ export type ParsedSitemap = {
183
+ url: string;
184
+ statusCode: number | null;
185
+ error: string | null;
186
+ contentType: string;
187
+ bytes: number;
188
+ gzipBroken: boolean;
189
+ malformed: boolean;
190
+ namespace: string | null;
191
+ isIndex: boolean;
192
+ nestedIndex: boolean;
193
+ urls: SitemapUrl[];
194
+ childSitemaps: string[];
195
+ relativeLocs: string[];
196
+ };
197
+ export type OriginProbe = {
198
+ https: boolean;
199
+ httpRedirectsToHttps: boolean | null;
200
+ wwwSplit: boolean | null;
201
+ wwwPreferred: string | null;
202
+ trailingSlashSplit: boolean | null;
203
+ randomSoft404: boolean;
204
+ randomSoft404Url: string | null;
205
+ randomSoft404Hash: string | null;
206
+ certValidTo: string | null;
207
+ certDaysRemaining: number | null;
208
+ certError: string | null;
209
+ alpn: string | null;
210
+ hsts: string | null;
211
+ };
212
+ export type CrawlCheckpoint = {
213
+ seen: string[];
214
+ queue: {
215
+ url: string;
216
+ depth: number;
217
+ }[];
218
+ pagesSeen: number;
219
+ };
220
+ export type CrawlResult = {
221
+ origin: string;
222
+ startUrl: string;
223
+ startedAt: string;
224
+ finishedAt: string;
225
+ pages: CrawledPage[];
226
+ robots: RobotsOutcome | null;
227
+ sitemaps: ParsedSitemap[];
228
+ originProbe: OriginProbe;
229
+ pagesSeen: number;
230
+ pagesChanged: number;
231
+ maxPages: number;
232
+ truncated: boolean;
233
+ aborted: boolean;
234
+ checkpoint: CrawlCheckpoint;
235
+ };
236
+ export type CrawlOptions = {
237
+ startUrl: string;
238
+ maxPages?: number | undefined;
239
+ concurrency?: number | undefined;
240
+ rps?: number | undefined;
241
+ timeoutMs?: number | undefined;
242
+ render?: boolean | undefined;
243
+ userAgent?: string | undefined;
244
+ followExternal?: boolean | undefined;
245
+ check304?: boolean | undefined;
246
+ signal?: AbortSignal | undefined;
247
+ resume?: CrawlCheckpoint | undefined;
248
+ onCheckpoint?: ((checkpoint: CrawlCheckpoint) => void) | undefined;
249
+ checkpointEvery?: number | undefined;
250
+ };
251
+ //# sourceMappingURL=types.d.ts.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"types.d.ts","sourceRoot":"","sources":["../src/types.ts"],"names":[],"mappings":"AAAA,MAAM,MAAM,WAAW,GAAG;IACxB,GAAG,EAAE,MAAM,CAAC;IACZ,MAAM,EAAE,MAAM,CAAC;IACf,QAAQ,EAAE,MAAM,GAAG,IAAI,CAAC;CACzB,CAAC;AAEF,MAAM,MAAM,aAAa,GAAG;IAC1B,IAAI,EAAE,MAAM,CAAC;IACb,MAAM,EAAE,MAAM,CAAC;IACf,UAAU,EAAE,MAAM,CAAC;IACnB,GAAG,EAAE,MAAM,EAAE,CAAC;IACd,MAAM,EAAE,MAAM,GAAG,IAAI,CAAC;IACtB,UAAU,EAAE,OAAO,CAAC;IACpB,KAAK,EAAE,OAAO,CAAC;IACf,WAAW,EAAE,OAAO,CAAC;IACrB,UAAU,EAAE,OAAO,CAAC;IACpB,QAAQ,EAAE,KAAK,GAAG,QAAQ,GAAG,QAAQ,GAAG,OAAO,GAAG,MAAM,GAAG,SAAS,CAAC;CACtE,CAAC;AAEF,MAAM,MAAM,cAAc,GAAG;IAC3B,GAAG,EAAE,MAAM,CAAC;IACZ,MAAM,EAAE,MAAM,CAAC;IACf,GAAG,EAAE,MAAM,GAAG,IAAI,CAAC;IACnB,UAAU,EAAE,OAAO,CAAC;IACpB,KAAK,EAAE,MAAM,GAAG,IAAI,CAAC;IACrB,MAAM,EAAE,MAAM,GAAG,IAAI,CAAC;IACtB,OAAO,EAAE,MAAM,GAAG,IAAI,CAAC;IACvB,MAAM,EAAE,MAAM,GAAG,IAAI,CAAC;IACtB,QAAQ,EAAE,MAAM,GAAG,IAAI,CAAC;CACzB,CAAC;AAEF,MAAM,MAAM,WAAW,GAAG;IACxB,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,MAAM,CAAC;IACb,MAAM,EAAE,MAAM,CAAC;IACf,MAAM,EAAE,OAAO,CAAC;IAChB,MAAM,EAAE,MAAM,GAAG,QAAQ,GAAG,MAAM,CAAC;CACpC,CAAC;AAEF,MAAM,MAAM,WAAW,GAAG;IACxB,GAAG,EAAE,MAAM,CAAC;IACZ,MAAM,EAAE,OAAO,CAAC;IAChB,KAAK,EAAE,MAAM,GAAG,IAAI,CAAC;IACrB,MAAM,EAAE,OAAO,CAAC;CACjB,CAAC;AAEF,MAAM,MAAM,WAAW,GAAG;IACxB,KAAK,EAAE,MAAM,GAAG,IAAI,CAAC;IACrB,MAAM,EAAE,MAAM,EAAE,CAAC;IACjB,WAAW,EAAE,OAAO,CAAC;IACrB,eAAe,EAAE,MAAM,GAAG,IAAI,CAAC;IAC/B,gBAAgB,EAAE,MAAM,EAAE,CAAC;IAC3B,cAAc,EAAE,OAAO,CAAC;IACxB,EAAE,EAAE,MAAM,EAAE,CAAC;IACb,EAAE,EAAE,MAAM,EAAE,CAAC;IACb,QAAQ,EAAE;QAAE,KAAK,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;IAC5C,aAAa,EAAE,MAAM,EAAE,CAAC;IACxB,eAAe,EAAE,MAAM,EAAE,CAAC;IAC1B,eAAe,EAAE,OAAO,CAAC;IACzB,oBAAoB,EAAE,OAAO,CAAC;IAC9B,eAAe,EAAE,OAAO,CAAC;IACzB,UAAU,EAAE,MAAM,EAAE,CAAC;IACrB,gBAAgB,EAAE,OAAO,CAAC;IAC1B,UAAU,EAAE,MAAM,GAAG,IAAI,CAAC;IAC1B,QAAQ,EAAE,MAAM,GAAG,IAAI,CAAC;IACxB,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;IACpB,OAAO,EAAE,MAAM,GAAG,IAAI,CAAC;IACvB,kBAAkB,EAAE,OAAO,CAAC;IAC5B,QAAQ,EAAE,WAAW,EAAE,CAAC;IACxB,MAAM,EAAE,WAAW,EAAE,CAAC;IACtB,SAAS,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAClC,WAAW,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IACpC,KAAK,EAAE,aAAa,EAAE,CAAC;IACvB,MAAM,EAAE,cAAc,EAAE,CAAC;IACzB,OAAO,EAAE;QACP,GAAG,EAAE,MAAM,GAAG,IAAI,CAAC;QACnB,MAAM,EAAE,MAAM,GAAG,IAAI,CAAC;QACtB,KAAK,EAAE,OAAO,CAAC;QACf,KAAK,EAAE,OAAO,CAAC;KAChB,EAAE,CAAC;IACJ,WAAW,EAAE;QAAE,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;QAAC,MAAM,EAAE,MAAM,GAAG,IAAI,CAAA;KAAE,EAAE,CAAC;IAC9D,SAAS,EAAE,MAAM,CAAC;IAClB,aAAa,EAAE,MAAM,CAAC;IACtB,QAAQ,EAAE,MAAM,CAAC;IACjB,OAAO,EAAE,MAAM,CAAC;IAChB,gBAAgB,EAAE,MAAM,CAAC;IACzB,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,oBAAoB,EAAE,OAAO,CAAC;IAC9B,YAAY,EAAE,OAAO,CAAC;IACtB,gBAAgB,EAAE,OAAO,CAAC;IAC1B,KAAK,EAAE;QAAE,MAAM,EAAE,MAAM,CAAC;QAAC,SAAS,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;IAC/C,WAAW,EAAE,MAAM,EAAE,CAAC;IACtB,YAAY,EAAE,MAAM,EAAE,CAAC;IACvB,kBAAkB,EAAE,MAAM,CAAC;IAC3B,OAAO,EAAE,OAAO,CAAC;IACjB,UAAU,EAAE,OAAO,CAAC;IACpB,aAAa,EAAE;QAAE,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;QAAC,IAAI,EAAE,MAAM,GAAG,IAAI,CAAA;KAAE,CAAC;IAC5D,QAAQ,EAAE,MAAM,GAAG,IAAI,CAAC;CACzB,CAAC;AAEF,MAAM,MAAM,WAAW,GAAG;IACxB,GAAG,EAAE,MAAM,CAAC;IACZ,QAAQ,EAAE,MAAM,CAAC;IACjB,UAAU,EAAE,MAAM,GAAG,IAAI,CAAC;IAC1B,KAAK,EAAE,MAAM,GAAG,IAAI,CAAC;IACrB,WAAW,EAAE,MAAM,CAAC;IACpB,OAAO,EAAE,MAAM,CAAC,MAAM,EAAE,MAAM,CAAC,CAAC;IAChC,MAAM,EAAE,MAAM,CAAC;IACf,OAAO,EAAE,MAAM,CAAC;IAChB,SAAS,EAAE,MAAM,CAAC;IAClB,YAAY,EAAE,MAAM,CAAC;IACrB,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,aAAa,EAAE,WAAW,EAAE,CAAC;IAC7B,YAAY,EAAE,OAAO,CAAC;IACtB,0BAA0B,EAAE,OAAO,CAAC;IACpC,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;IACpB,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;IAC5B,OAAO,EAAE,WAAW,GAAG,IAAI,CAAC;IAC5B,eAAe,EAAE,WAAW,GAAG,IAAI,CAAC;IACpC,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,OAAO,EAAE,MAAM,GAAG,IAAI,CAAC;IACvB,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;IACpB,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;IAC5B,WAAW,EAAE,OAAO,CAAC;IACrB,WAAW,EAAE,OAAO,GAAG,IAAI,CAAC;IAC5B,SAAS,EAAE,OAAO,CAAC;IACnB,aAAa,EAAE,OAAO,CAAC;IACvB,KAAK,EAAE,MAAM,CAAC;IACd,WAAW,EAAE,MAAM,CAAC;IACpB,YAAY,EAAE,MAAM,CAAC;IACrB,iBAAiB,EAAE,MAAM,CAAC;IAC1B,UAAU,EAAE,OAAO,CAAC;IACpB,SAAS,EAAE,MAAM,CAAC;IAClB,YAAY,EAAE,QAAQ,GAAG,OAAO,GAAG,QAAQ,GAAG,SAAS,CAAC;IACxD,QAAQ,EAAE,OAAO,CAAC;IAClB,iBAAiB,EAAE,MAAM,CAAC;IAC1B,WAAW,EAAE,MAAM,CAAC;CACrB,CAAC;AAEF,MAAM,MAAM,aAAa,GAAG;IAC1B,UAAU,EAAE,MAAM,GAAG,IAAI,CAAC;IAC1B,GAAG,EAAE,MAAM,CAAC;IACZ,WAAW,EAAE,MAAM,CAAC;IACpB,KAAK,EAAE,MAAM,CAAC;IACd,YAAY,EAAE,MAAM,CAAC;IACrB,KAAK,EAAE,MAAM,GAAG,IAAI,CAAC;IACrB,MAAM,EAAE,WAAW,EAAE,CAAC;IACtB,QAAQ,EAAE,MAAM,EAAE,CAAC;IACnB,iBAAiB,EAAE,MAAM,EAAE,CAAC;IAC5B,mBAAmB,EAAE,OAAO,CAAC;IAC7B,aAAa,EAAE,OAAO,CAAC;IACvB,UAAU,EAAE,MAAM,GAAG,IAAI,CAAC;IAC1B,WAAW,EAAE,OAAO,CAAC;IACrB,IAAI,EAAE,MAAM,CAAC;IACb,IAAI,EAAE,WAAW,GAAG,cAAc,GAAG,QAAQ,CAAC;CAC/C,CAAC;AAEF,MAAM,MAAM,WAAW,GAAG;IACxB,MAAM,EAAE,MAAM,EAAE,CAAC;IACjB,KAAK,EAAE;QAAE,IAAI,EAAE,OAAO,GAAG,UAAU,CAAC;QAAC,OAAO,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;IACzD,UAAU,EAAE,MAAM,GAAG,IAAI,CAAC;CAC3B,CAAC;AAEF,MAAM,MAAM,UAAU,GAAG;IACvB,GAAG,EAAE,MAAM,CAAC;IACZ,OAAO,EAAE,MAAM,GAAG,IAAI,CAAC;IACvB,UAAU,EAAE,MAAM,GAAG,IAAI,CAAC;IAC1B,QAAQ,EAAE,MAAM,GAAG,IAAI,CAAC;IACxB,MAAM,EAAE,MAAM,EAAE,CAAC;IACjB,UAAU,EAAE,OAAO,GAAG,IAAI,CAAC;IAC3B,QAAQ,EAAE;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;CAC5C,CAAC;AAEF,MAAM,MAAM,aAAa,GAAG;IAC1B,GAAG,EAAE,MAAM,CAAC;IACZ,UAAU,EAAE,MAAM,GAAG,IAAI,CAAC;IAC1B,KAAK,EAAE,MAAM,GAAG,IAAI,CAAC;IACrB,WAAW,EAAE,MAAM,CAAC;IACpB,KAAK,EAAE,MAAM,CAAC;IACd,UAAU,EAAE,OAAO,CAAC;IACpB,SAAS,EAAE,OAAO,CAAC;IACnB,SAAS,EAAE,MAAM,GAAG,IAAI,CAAC;IACzB,OAAO,EAAE,OAAO,CAAC;IACjB,WAAW,EAAE,OAAO,CAAC;IACrB,IAAI,EAAE,UAAU,EAAE,CAAC;IACnB,aAAa,EAAE,MAAM,EAAE,CAAC;IACxB,YAAY,EAAE,MAAM,EAAE,CAAC;CACxB,CAAC;AAEF,MAAM,MAAM,WAAW,GAAG;IACxB,KAAK,EAAE,OAAO,CAAC;IACf,oBAAoB,EAAE,OAAO,GAAG,IAAI,CAAC;IACrC,QAAQ,EAAE,OAAO,GAAG,IAAI,CAAC;IACzB,YAAY,EAAE,MAAM,GAAG,IAAI,CAAC;IAC5B,kBAAkB,EAAE,OAAO,GAAG,IAAI,CAAC;IACnC,aAAa,EAAE,OAAO,CAAC;IACvB,gBAAgB,EAAE,MAAM,GAAG,IAAI,CAAC;IAChC,iBAAiB,EAAE,MAAM,GAAG,IAAI,CAAC;IACjC,WAAW,EAAE,MAAM,GAAG,IAAI,CAAC;IAC3B,iBAAiB,EAAE,MAAM,GAAG,IAAI,CAAC;IACjC,SAAS,EAAE,MAAM,GAAG,IAAI,CAAC;IACzB,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;IACpB,IAAI,EAAE,MAAM,GAAG,IAAI,CAAC;CACrB,CAAC;AAEF,MAAM,MAAM,eAAe,GAAG;IAC5B,IAAI,EAAE,MAAM,EAAE,CAAC;IACf,KAAK,EAAE;QAAE,GAAG,EAAE,MAAM,CAAC;QAAC,KAAK,EAAE,MAAM,CAAA;KAAE,EAAE,CAAC;IACxC,SAAS,EAAE,MAAM,CAAC;CACnB,CAAC;AAEF,MAAM,MAAM,WAAW,GAAG;IACxB,MAAM,EAAE,MAAM,CAAC;IACf,QAAQ,EAAE,MAAM,CAAC;IACjB,SAAS,EAAE,MAAM,CAAC;IAClB,UAAU,EAAE,MAAM,CAAC;IACnB,KAAK,EAAE,WAAW,EAAE,CAAC;IACrB,MAAM,EAAE,aAAa,GAAG,IAAI,CAAC;IAC7B,QAAQ,EAAE,aAAa,EAAE,CAAC;IAC1B,WAAW,EAAE,WAAW,CAAC;IACzB,SAAS,EAAE,MAAM,CAAC;IAClB,YAAY,EAAE,MAAM,CAAC;IACrB,QAAQ,EAAE,MAAM,CAAC;IACjB,SAAS,EAAE,OAAO,CAAC;IACnB,OAAO,EAAE,OAAO,CAAC;IACjB,UAAU,EAAE,eAAe,CAAC;CAC7B,CAAC;AAEF,MAAM,MAAM,YAAY,GAAG;IACzB,QAAQ,EAAE,MAAM,CAAC;IACjB,QAAQ,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC9B,WAAW,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IACjC,GAAG,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IACzB,SAAS,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC/B,MAAM,CAAC,EAAE,OAAO,GAAG,SAAS,CAAC;IAC7B,SAAS,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;IAC/B,cAAc,CAAC,EAAE,OAAO,GAAG,SAAS,CAAC;IACrC,QAAQ,CAAC,EAAE,OAAO,GAAG,SAAS,CAAC;IAC/B,MAAM,CAAC,EAAE,WAAW,GAAG,SAAS,CAAC;IACjC,MAAM,CAAC,EAAE,eAAe,GAAG,SAAS,CAAC;IACrC,YAAY,CAAC,EAAE,CAAC,CAAC,UAAU,EAAE,eAAe,KAAK,IAAI,CAAC,GAAG,SAAS,CAAC;IACnE,eAAe,CAAC,EAAE,MAAM,GAAG,SAAS,CAAC;CACtC,CAAC"}
package/dist/types.js ADDED
@@ -0,0 +1,2 @@
1
+ export {};
2
+ //# sourceMappingURL=types.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"types.js","sourceRoot":"","sources":["../src/types.ts"],"names":[],"mappings":""}
package/dist/ua.d.ts ADDED
@@ -0,0 +1,3 @@
1
+ export declare const SEAN_UA = "Mozilla/5.0 (compatible; SeanBot/0.0; +https://github.com/seziro-team/agentsean) undici/8";
2
+ export declare const GOOGLEBOT_SMARTPHONE_UA = "Mozilla/5.0 (Linux; Android 6.0.1; Nexus 5X Build/MMB29P) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Mobile Safari/537.36 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)";
3
+ //# sourceMappingURL=ua.d.ts.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"ua.d.ts","sourceRoot":"","sources":["../src/ua.ts"],"names":[],"mappings":"AAAA,eAAO,MAAM,OAAO,8FACyE,CAAC;AAE9F,eAAO,MAAM,uBAAuB,wMACmK,CAAC"}
package/dist/ua.js ADDED
@@ -0,0 +1,3 @@
1
+ export const SEAN_UA = "Mozilla/5.0 (compatible; SeanBot/0.0; +https://github.com/seziro-team/agentsean) undici/8";
2
+ export const GOOGLEBOT_SMARTPHONE_UA = "Mozilla/5.0 (Linux; Android 6.0.1; Nexus 5X Build/MMB29P) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Mobile Safari/537.36 (compatible; Googlebot/2.1; +http://www.google.com/bot.html)";
3
+ //# sourceMappingURL=ua.js.map
package/dist/ua.js.map ADDED
@@ -0,0 +1 @@
1
+ {"version":3,"file":"ua.js","sourceRoot":"","sources":["../src/ua.ts"],"names":[],"mappings":"AAAA,MAAM,CAAC,MAAM,OAAO,GAClB,2FAA2F,CAAC;AAE9F,MAAM,CAAC,MAAM,uBAAuB,GAClC,qMAAqM,CAAC"}
package/dist/url.d.ts ADDED
@@ -0,0 +1,11 @@
1
+ export declare function originOf(url: string): string;
2
+ export declare function registrableHost(hostname: string): string;
3
+ export declare function sameOrigin(a: string, b: string): boolean;
4
+ export declare function absolutize(href: string, base: string): string | null;
5
+ export declare function normalizeUrl(url: string): string;
6
+ export declare function stripTrackingParams(url: string): string;
7
+ export declare function hasTrackingParams(url: string): boolean;
8
+ export declare function isInternalUrl(url: string, origin: string): boolean;
9
+ export declare function pathOf(url: string): string;
10
+ export declare function templateKey(url: string): string;
11
+ //# sourceMappingURL=url.d.ts.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"url.d.ts","sourceRoot":"","sources":["../src/url.ts"],"names":[],"mappings":"AAyBA,wBAAgB,QAAQ,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,CAG5C;AAED,wBAAgB,eAAe,CAAC,QAAQ,EAAE,MAAM,GAAG,MAAM,CAIxD;AAED,wBAAgB,UAAU,CAAC,CAAC,EAAE,MAAM,EAAE,CAAC,EAAE,MAAM,GAAG,OAAO,CAMxD;AAED,wBAAgB,UAAU,CAAC,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,MAAM,GAAG,MAAM,GAAG,IAAI,CAMpE;AAED,wBAAgB,YAAY,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,CAKhD;AAED,wBAAgB,mBAAmB,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,CAOvD;AAED,wBAAgB,iBAAiB,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAUtD;AAED,wBAAgB,aAAa,CAAC,GAAG,EAAE,MAAM,EAAE,MAAM,EAAE,MAAM,GAAG,OAAO,CAYlE;AAED,wBAAgB,MAAM,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,CAO1C;AAED,wBAAgB,WAAW,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,CAW/C"}
package/dist/url.js ADDED
@@ -0,0 +1,114 @@
1
+ const TRACKING_PARAMS = new Set([
2
+ "utm_source",
3
+ "utm_medium",
4
+ "utm_campaign",
5
+ "utm_term",
6
+ "utm_content",
7
+ "utm_id",
8
+ "gclid",
9
+ "gbraid",
10
+ "wbraid",
11
+ "fbclid",
12
+ "msclkid",
13
+ "ttclid",
14
+ "twclid",
15
+ "li_fat_id",
16
+ "igshid",
17
+ "mc_cid",
18
+ "mc_eid",
19
+ "_ga",
20
+ "_gl",
21
+ "yclid",
22
+ "ref",
23
+ "referrer",
24
+ ]);
25
+ export function originOf(url) {
26
+ const u = new URL(url);
27
+ return `${u.protocol}//${u.host}`;
28
+ }
29
+ export function registrableHost(hostname) {
30
+ const parts = hostname.replace(/\.$/, "").toLowerCase().split(".");
31
+ if (parts.length <= 2)
32
+ return parts.join(".");
33
+ return parts.slice(-2).join(".");
34
+ }
35
+ export function sameOrigin(a, b) {
36
+ try {
37
+ return new URL(a).origin === new URL(b).origin;
38
+ }
39
+ catch {
40
+ return false;
41
+ }
42
+ }
43
+ export function absolutize(href, base) {
44
+ try {
45
+ return new URL(href, base).href;
46
+ }
47
+ catch {
48
+ return null;
49
+ }
50
+ }
51
+ export function normalizeUrl(url) {
52
+ const u = new URL(url);
53
+ u.hash = "";
54
+ if (u.pathname === "")
55
+ u.pathname = "/";
56
+ return u.href;
57
+ }
58
+ export function stripTrackingParams(url) {
59
+ const u = new URL(url);
60
+ const keys = Array.from(u.searchParams.keys());
61
+ for (const key of keys) {
62
+ if (TRACKING_PARAMS.has(key.toLowerCase()))
63
+ u.searchParams.delete(key);
64
+ }
65
+ return u.href;
66
+ }
67
+ export function hasTrackingParams(url) {
68
+ try {
69
+ const u = new URL(url);
70
+ for (const key of u.searchParams.keys()) {
71
+ if (TRACKING_PARAMS.has(key.toLowerCase()))
72
+ return true;
73
+ }
74
+ }
75
+ catch {
76
+ return false;
77
+ }
78
+ return false;
79
+ }
80
+ export function isInternalUrl(url, origin) {
81
+ try {
82
+ const u = new URL(url);
83
+ const o = new URL(origin);
84
+ return (u.hostname === o.hostname ||
85
+ u.hostname === `www.${o.hostname}` ||
86
+ o.hostname === `www.${u.hostname}`);
87
+ }
88
+ catch {
89
+ return false;
90
+ }
91
+ }
92
+ export function pathOf(url) {
93
+ try {
94
+ const u = new URL(url);
95
+ return `${u.pathname}${u.search}`;
96
+ }
97
+ catch {
98
+ return url;
99
+ }
100
+ }
101
+ export function templateKey(url) {
102
+ try {
103
+ const u = new URL(url);
104
+ const segs = u.pathname
105
+ .split("/")
106
+ .filter(Boolean)
107
+ .map((s) => (/^\d+$/.test(s) || /^[0-9a-f-]{8,}$/i.test(s) ? ":id" : s));
108
+ return `${segs.length}:${segs.join("/")}`;
109
+ }
110
+ catch {
111
+ return "0:";
112
+ }
113
+ }
114
+ //# sourceMappingURL=url.js.map
@@ -0,0 +1 @@
1
+ {"version":3,"file":"url.js","sourceRoot":"","sources":["../src/url.ts"],"names":[],"mappings":"AAAA,MAAM,eAAe,GAAG,IAAI,GAAG,CAAC;IAC9B,YAAY;IACZ,YAAY;IACZ,cAAc;IACd,UAAU;IACV,aAAa;IACb,QAAQ;IACR,OAAO;IACP,QAAQ;IACR,QAAQ;IACR,QAAQ;IACR,SAAS;IACT,QAAQ;IACR,QAAQ;IACR,WAAW;IACX,QAAQ;IACR,QAAQ;IACR,QAAQ;IACR,KAAK;IACL,KAAK;IACL,OAAO;IACP,KAAK;IACL,UAAU;CACX,CAAC,CAAC;AAEH,MAAM,UAAU,QAAQ,CAAC,GAAW;IAClC,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;IACvB,OAAO,GAAG,CAAC,CAAC,QAAQ,KAAK,CAAC,CAAC,IAAI,EAAE,CAAC;AACpC,CAAC;AAED,MAAM,UAAU,eAAe,CAAC,QAAgB;IAC9C,MAAM,KAAK,GAAG,QAAQ,CAAC,OAAO,CAAC,KAAK,EAAE,EAAE,CAAC,CAAC,WAAW,EAAE,CAAC,KAAK,CAAC,GAAG,CAAC,CAAC;IACnE,IAAI,KAAK,CAAC,MAAM,IAAI,CAAC;QAAE,OAAO,KAAK,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;IAC9C,OAAO,KAAK,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;AACnC,CAAC;AAED,MAAM,UAAU,UAAU,CAAC,CAAS,EAAE,CAAS;IAC7C,IAAI,CAAC;QACH,OAAO,IAAI,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,KAAK,IAAI,GAAG,CAAC,CAAC,CAAC,CAAC,MAAM,CAAC;IACjD,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC;AAED,MAAM,UAAU,UAAU,CAAC,IAAY,EAAE,IAAY;IACnD,IAAI,CAAC;QACH,OAAO,IAAI,GAAG,CAAC,IAAI,EAAE,IAAI,CAAC,CAAC,IAAI,CAAC;IAClC,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,IAAI,CAAC;IACd,CAAC;AACH,CAAC;AAED,MAAM,UAAU,YAAY,CAAC,GAAW;IACtC,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;IACvB,CAAC,CAAC,IAAI,GAAG,EAAE,CAAC;IACZ,IAAI,CAAC,CAAC,QAAQ,KAAK,EAAE;QAAE,CAAC,CAAC,QAAQ,GAAG,GAAG,CAAC;IACxC,OAAO,CAAC,CAAC,IAAI,CAAC;AAChB,CAAC;AAED,MAAM,UAAU,mBAAmB,CAAC,GAAW;IAC7C,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;IACvB,MAAM,IAAI,GAAG,KAAK,CAAC,IAAI,CAAC,CAAC,CAAC,YAAY,CAAC,IAAI,EAAE,CAAC,CAAC;IAC/C,KAAK,MAAM,GAAG,IAAI,IAAI,EAAE,CAAC;QACvB,IAAI,eAAe,CAAC,GAAG,CAAC,GAAG,CAAC,WAAW,EAAE,CAAC;YAAE,CAAC,CAAC,YAAY,CAAC,MAAM,CAAC,GAAG,CAAC,CAAC;IACzE,CAAC;IACD,OAAO,CAAC,CAAC,IAAI,CAAC;AAChB,CAAC;AAED,MAAM,UAAU,iBAAiB,CAAC,GAAW;IAC3C,IAAI,CAAC;QACH,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;QACvB,KAAK,MAAM,GAAG,IAAI,CAAC,CAAC,YAAY,CAAC,IAAI,EAAE,EAAE,CAAC;YACxC,IAAI,eAAe,CAAC,GAAG,CAAC,GAAG,CAAC,WAAW,EAAE,CAAC;gBAAE,OAAO,IAAI,CAAC;QAC1D,CAAC;IACH,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;IACD,OAAO,KAAK,CAAC;AACf,CAAC;AAED,MAAM,UAAU,aAAa,CAAC,GAAW,EAAE,MAAc;IACvD,IAAI,CAAC;QACH,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;QACvB,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,MAAM,CAAC,CAAC;QAC1B,OAAO,CACL,CAAC,CAAC,QAAQ,KAAK,CAAC,CAAC,QAAQ;YACzB,CAAC,CAAC,QAAQ,KAAK,OAAO,CAAC,CAAC,QAAQ,EAAE;YAClC,CAAC,CAAC,QAAQ,KAAK,OAAO,CAAC,CAAC,QAAQ,EAAE,CACnC,CAAC;IACJ,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,KAAK,CAAC;IACf,CAAC;AACH,CAAC;AAED,MAAM,UAAU,MAAM,CAAC,GAAW;IAChC,IAAI,CAAC;QACH,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;QACvB,OAAO,GAAG,CAAC,CAAC,QAAQ,GAAG,CAAC,CAAC,MAAM,EAAE,CAAC;IACpC,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,GAAG,CAAC;IACb,CAAC;AACH,CAAC;AAED,MAAM,UAAU,WAAW,CAAC,GAAW;IACrC,IAAI,CAAC;QACH,MAAM,CAAC,GAAG,IAAI,GAAG,CAAC,GAAG,CAAC,CAAC;QACvB,MAAM,IAAI,GAAG,CAAC,CAAC,QAAQ;aACpB,KAAK,CAAC,GAAG,CAAC;aACV,MAAM,CAAC,OAAO,CAAC;aACf,GAAG,CAAC,CAAC,CAAC,EAAE,EAAE,CAAC,CAAC,OAAO,CAAC,IAAI,CAAC,CAAC,CAAC,IAAI,kBAAkB,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC,CAAC;QAC3E,OAAO,GAAG,IAAI,CAAC,MAAM,IAAI,IAAI,CAAC,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC;IAC5C,CAAC;IAAC,MAAM,CAAC;QACP,OAAO,IAAI,CAAC;IACd,CAAC;AACH,CAAC"}
package/package.json ADDED
@@ -0,0 +1,60 @@
1
+ {
2
+ "name": "@agentsean/crawler",
3
+ "version": "1.0.0",
4
+ "license": "AGPL-3.0-only",
5
+ "type": "module",
6
+ "exports": {
7
+ ".": {
8
+ "types": "./dist/index.d.ts",
9
+ "import": "./dist/index.js"
10
+ }
11
+ },
12
+ "files": [
13
+ "dist"
14
+ ],
15
+ "engines": {
16
+ "node": ">=22.19.0"
17
+ },
18
+ "dependencies": {
19
+ "drizzle-orm": "0.45.2",
20
+ "cheerio": "1.2.0",
21
+ "htmlparser2": "12.0.0",
22
+ "linkedom": "0.18.13",
23
+ "parse5": "8.0.1",
24
+ "saxes": "6.0.0",
25
+ "undici": "8.10.1",
26
+ "@agentsean/db": "1.0.0"
27
+ },
28
+ "optionalDependencies": {
29
+ "playwright": "1.62.1"
30
+ },
31
+ "devDependencies": {
32
+ "typescript": "~5.9.3",
33
+ "vitest": "^4.1.11"
34
+ },
35
+ "publishConfig": {
36
+ "access": "public"
37
+ },
38
+ "repository": {
39
+ "type": "git",
40
+ "url": "git+https://github.com/seziro-team/agentsean.git",
41
+ "directory": "packages/crawler"
42
+ },
43
+ "homepage": "https://github.com/seziro-team/agentsean#readme",
44
+ "bugs": {
45
+ "url": "https://github.com/seziro-team/agentsean/issues"
46
+ },
47
+ "author": "Seziro (https://seziro.com)",
48
+ "description": "Agent Sean — crawler module",
49
+ "keywords": [
50
+ "seo",
51
+ "agent",
52
+ "agentsean",
53
+ "seo-automation"
54
+ ],
55
+ "scripts": {
56
+ "build": "tsc -p tsconfig.json",
57
+ "typecheck": "tsc -p tsconfig.json --noEmit",
58
+ "test": "vitest run --config ../../vitest.config.ts"
59
+ }
60
+ }