1 | import { Matcher, MatcherConfig } from "./matcher";
|
2 | import { StripPrefixConfigObj } from "../autolinker";
|
3 | import { Match } from "../match/match";
|
4 | /**
|
5 | * @class Autolinker.matcher.Url
|
6 | * @extends Autolinker.matcher.Matcher
|
7 | *
|
8 | * Matcher to find URL matches in an input string.
|
9 | *
|
10 | * See this class's superclass ({@link Autolinker.matcher.Matcher}) for more details.
|
11 | */
|
12 | export declare class UrlMatcher extends Matcher {
|
13 | /**
|
14 | * @cfg {Object} stripPrefix (required)
|
15 | *
|
16 | * The Object form of {@link Autolinker#cfg-stripPrefix}.
|
17 | */
|
18 | protected stripPrefix: Required<StripPrefixConfigObj>;
|
19 | /**
|
20 | * @cfg {Boolean} stripTrailingSlash (required)
|
21 | * @inheritdoc Autolinker#stripTrailingSlash
|
22 | */
|
23 | protected stripTrailingSlash: boolean;
|
24 | /**
|
25 | * @cfg {Boolean} decodePercentEncoding (required)
|
26 | * @inheritdoc Autolinker#decodePercentEncoding
|
27 | */
|
28 | protected decodePercentEncoding: boolean;
|
29 | /**
|
30 | * @protected
|
31 | * @property {RegExp} matcherRegex
|
32 | *
|
33 | * The regular expression to match URLs with an optional scheme, port
|
34 | * number, path, query string, and hash anchor.
|
35 | *
|
36 | * Example matches:
|
37 | *
|
38 | * http://google.com
|
39 | * www.google.com
|
40 | * google.com/path/to/file?q1=1&q2=2#myAnchor
|
41 | *
|
42 | *
|
43 | * This regular expression will have the following capturing groups:
|
44 | *
|
45 | * 1. Group that matches a scheme-prefixed URL (i.e. 'http://google.com').
|
46 | * This is used to match scheme URLs with just a single word, such as
|
47 | * 'http://localhost', where we won't double check that the domain name
|
48 | * has at least one dot ('.') in it.
|
49 | * 2. Group that matches a 'www.' prefixed URL. This is only matched if the
|
50 | * 'www.' text was not prefixed by a scheme (i.e.: not prefixed by
|
51 | * 'http://', 'ftp:', etc.)
|
52 | * 3. A protocol-relative ('//') match for the case of a 'www.' prefixed
|
53 | * URL. Will be an empty string if it is not a protocol-relative match.
|
54 | * We need to know the character before the '//' in order to determine
|
55 | * if it is a valid match or the // was in a string we don't want to
|
56 | * auto-link.
|
57 | * 4. Group that matches a known TLD (top level domain), when a scheme
|
58 | * or 'www.'-prefixed domain is not matched.
|
59 | * 5. A protocol-relative ('//') match for the case of a known TLD prefixed
|
60 | * URL. Will be an empty string if it is not a protocol-relative match.
|
61 | * See #3 for more info.
|
62 | */
|
63 | protected matcherRegex: RegExp;
|
64 | /**
|
65 | * A regular expression to use to check the character before a protocol-relative
|
66 | * URL match. We don't want to match a protocol-relative URL if it is part
|
67 | * of another word.
|
68 | *
|
69 | * For example, we want to match something like "Go to: //google.com",
|
70 | * but we don't want to match something like "abc//google.com"
|
71 | *
|
72 | * This regular expression is used to test the character before the '//'.
|
73 | *
|
74 | * @protected
|
75 | * @type {RegExp} wordCharRegExp
|
76 | */
|
77 | protected wordCharRegExp: RegExp;
|
78 | /**
|
79 | * @method constructor
|
80 | * @param {Object} cfg The configuration properties for the Match instance,
|
81 | * specified in an Object (map).
|
82 | */
|
83 | constructor(cfg: UrlMatcherConfig);
|
84 | /**
|
85 | * @inheritdoc
|
86 | */
|
87 | parseMatches(text: string): Match[];
|
88 | /**
|
89 | * Determines if a match found has an unmatched closing parenthesis,
|
90 | * square bracket or curly bracket. If so, the symbol will be removed
|
91 | * from the match itself, and appended after the generated anchor tag.
|
92 | *
|
93 | * A match may have an extra closing parenthesis at the end of the match
|
94 | * because the regular expression must include parenthesis for URLs such as
|
95 | * "wikipedia.com/something_(disambiguation)", which should be auto-linked.
|
96 | *
|
97 | * However, an extra parenthesis *will* be included when the URL itself is
|
98 | * wrapped in parenthesis, such as in the case of:
|
99 | * "(wikipedia.com/something_(disambiguation))"
|
100 | * In this case, the last closing parenthesis should *not* be part of the
|
101 | * URL itself, and this method will return `true`.
|
102 | *
|
103 | * For square brackets in URLs such as in PHP arrays, the same behavior as
|
104 | * parenthesis discussed above should happen:
|
105 | * "[http://www.example.com/foo.php?bar[]=1&bar[]=2&bar[]=3]"
|
106 | * The closing square bracket should not be part of the URL itself, and this
|
107 | * method will return `true`.
|
108 | *
|
109 | * @protected
|
110 | * @param {String} matchStr The full match string from the { #matcherRegex}.
|
111 | * Boolean} `true` if there is an unbalanced closing parenthesis or
{ |
112 | * square bracket at the end of the `matchStr`, `false` otherwise.
|
113 | */
|
114 | protected matchHasUnbalancedClosingParen(matchStr: string): boolean;
|
115 | /**
|
116 | * Determine if there's an invalid character after the TLD in a URL. Valid
|
117 | * characters after TLD are ':/?#'. Exclude scheme matched URLs from this
|
118 | * check.
|
119 | *
|
120 | * @protected
|
121 | * @param {String} urlMatch The matched URL, if there was one. Will be an
|
122 | * empty string if the match is not a URL match.
|
123 | * @param {String} schemeUrlMatch The match URL string for a scheme
|
124 | * match. Ex: 'http://yahoo.com'. This is used to match something like
|
125 | * 'http://localhost', where we won't double check that the domain name
|
126 | * has at least one '.' in it.
|
127 | * @return {Number} the position where the invalid character was found. If
|
128 | * no such character was found, returns -1
|
129 | */
|
130 | protected matchHasInvalidCharAfterTld(urlMatch: string, schemeUrlMatch: string): number;
|
131 | }
|
132 | export interface UrlMatcherConfig extends MatcherConfig {
|
133 | stripPrefix: Required<StripPrefixConfigObj>;
|
134 | stripTrailingSlash: boolean;
|
135 | decodePercentEncoding: boolean;
|
136 | }
|