.net - Prepend a URL segment to a relative / relative to server URL -

i have requirement prepend url segment relative, or relative-to-server urls within html document (eg href or src attributes) on server-side application

i'm working in .net environment, , unfortunately there no base class libraries loading html string dom , manipulating (i not have luxury of being able introduce third-party library @ point), seems candidate regex replace, i'm little weak on

i need regex can handle these 2 cases:

relative server url:

href="/controller/action" -> href="/mypathsegment/controller/action"

relative url:

href="image/logo.gif" -> href="/mypathsegments/image/logo.gif"

don't worry mypathsegment ..i have logic elsewhere can figure out levels of path segments relative urls, want focus on regex problem

i figure need regex can match on src="..." or href="/..." pattern , insert string @ beginning after first opening double quote not experienced enough regexs figure out.

you want find url, assert url not contain /mypathsegment/, , replace same property (href or src), , same end-of-url, in middle, put path segment:

(?<tag>(?:href|src)=")(?(?=/mypathsegment/)(?!)|/?(?<url>[^"]*")) 

replace with:

${tag}/mypathsegment/${url} 

in vb.net:

dim strpathprefix string = "/mypathsegment/" dim strtest string = regex.replace("<a href=""/controller/action"" property=""something"">hello world</a>", "(?<tag>(?:href|src)="")(?(?=" & strpathprefix & ")(?!))/?(?<url>[^""]*"")", "${tag}/mypathsegment/${url}")  strtest = <a href="/mypathsegment/controller/action" property="something">hello world</a> 

note in vb, had double quotes.

with use, whether or not have slash irrelevant. you'd have make sure strpathprepend ends "/"

test on things , let me know if hit bumps might require tweaking.

(?<tag>... 

the data following should captured capture group named "tag".

(?:href|src) 

match either "href" or "src". "?:" means not capture results particular sub-expression, though results capture parent sub-expression "tag"

=") 

capture equal sign , quote symbol literally, , close capture group "tag"

(?(...)...|...) 

example conditional statement. if in small sub-expression, left-side of alternation "|", else on right-side of alternation.

(?=...) 

example lookaround. in case, lookaround "look-ahead". means: check if text following current position equals something, don't advance cursor.

(?(?=/mypathsegment/)... 

so, put our 2 examples together, , conditional says... if, after collecting href=", find /mypathsegment/, then...

(?!) 

this lookaround. negative lookahead. there nothing in it. (?!...). since negative lookahead , it's empty, automatically fails regex statement. want fail if sees path segment in there.

|/? 

else, if /mypathsegment/ not found... optionally find slash, isn't in capture group.

(?<url>... 

also in else statement... after optional slash, build our second capture group named "url"

[^"]*" 

capture not quote... multiple times (optional, change * + if want require something), after catching not quote... capture quote "url" capture group also.

)) 

first closes url capture group, closes conditional group.