Is something like this what you are looking for?
julia> matchall(r"(\d+?.+?\.ts)", s)
3-element Array{SubString{UTF8String},1}:
"00001.ts"
"00002.ts"
"00xyz.ts"
El jueves, 24 de diciembre de 2015, 12:30:52 (UTC-6), Douglas Bates
escribió:
>
> Short version:
>
> I have a string that contains several instances of names of the form
> 00001.ts, 00002.ts, ...., 00xyz.ts and I want to find the last match.
> That is, I want to find "00xyz.ts" or, alternatively, find all such names
> in sequence..
>
> Longer version:
>
> These are file names of a series of transport stream files containing
> audio, video, close captions, etc. The device generating them, a Tablo
> over-the-air video recorder, http://tabotv.com, breaks a recording into
> many small segments with these names. A program can query the device at a
> particular URL and get a listing of the directory containing these
> segments, returned as XHTML. I want all the file names in sequence so that
> I can download each of these files in sequence and create a single file by
> appending them.
>
> The names each occur multiple times in the string but each one only once
> in the form that will match r">(\d+\.ts)<".
>
> I can think of two ways of getting these names. One is to parse the
> string as XHTML and walk through the object to find these names. The other
> is to match the regular expression in the string, extract the "captures"
> field, match again starting at the current offset + 1, and continue until
> there are no further matches.
>
> The XML approach is more elegant but not especially easy. The regular
> expression matching is reasonably straightforward to implement but more
> fragile.
>
> Am I missing an elegant, robust approach here?
>
