StringRegExp and repeating a Capturing Group {x} times

[DobraGolonka]

Hi

I have the following sample line of text from a TV episode guide, which is also contained in the attached file:

<tr style="background: url(/_layout_v3/misc/tabletopicbg.jpg) repeat;" ><td class='b2'><a class='wlink' href='/24/episodes/613/01x24'>24 :01x24 - 11:00 P.M.-12:00 A.M.</a> (May/21/2002)<tr style='background: #EEEEEE; color: #000000;' ><td class='b2'><font color='black'>Sherry Palmer makes a move in her discontent over David's faked death. Jack offers himself in exchange for his daughter's life and squares off against the Drazens. Nina, the second mole, makes a hasty escape from CTU headquarters that is delayed when Teri intrudes. David Palmer makes a final decision concerning Sherry.</font></td></tr><tr bgcolor='#FFFFFF' id="brow"><td class='b2'><u><b>Guest Stars: </b></u><a href='/person/id-1046/Xander+Berkeley'>Xander Berkeley</a> as George Mason, <a href='/person/id-10013/Zeljko+Ivanek'>Zeljko Ivanek</a> as Andre Drazen, <a href='/person/id-1154/Terrell+Tilford'>Terrell Tilford</a> as Paul Wilson, <a href='/person/id-49206/Karina+Arroyave'>Karina Arroyave</a> as Jamey Farrell, <a href='/person/id-1044/Penny+Johnson+Jerald'>Penny Johnson Jerald</a> as Sherry Palmer, <a href='/person/id-1139/Dennis+Hopper'>Dennis Hopper</a> as Victor Drazen, <a href='/person/id-51174/Carlos+Bernard'>Carlos Bernard (2)</a> as Tony Almeida, <a href='/person/id-23977/Kevin+Chapman'>Kevin Chapman</a> as Coast Guard Officer, <a href='/person/id-28007/Jude+Ciccolella'>Jude Ciccolella</a> as Mike Novick, | <u><b>Co-Guest Stars: </b></u><a href='/person/id-1153/Tico+Wells'>Tico Wells</a> as Karris, <a href='/person/id-1152/Jane+Yamamoto'>Jane Yamamoto</a> as Field Reporter, <a href='/person/id-1144/Endre+Hules'>Endre Hules</a> as Serge, <a href='/person/id-52211/Reynaldo+Gallegos'>Reynaldo Gallegos (1)</a> as Sergeant Devlin</td></Tr><tr bgcolor='#E7E7CB' id="brow"><td class='b2'><u><b>Director: </b></u><a href='/person/id-1048/Stephen+Hopkins'>Stephen Hopkins</a><br><u><b>Writer: </b></u><a href='/person/id-1049/Joel+Surnow'>Joel Surnow</a>, <a href='/person/id-1058/Michael+Loceff'>Michael Loceff</a></td></Tr></table></td></tr><tr><td height='5'>&nbsp;</td></tr></table></td>

Within this text, I need to extract the following fields: ID, Series, Episode, Title, Airdate, Plot, Cast and Crew.

The following RegEx successfully extracts the fields ID-Plot with no problems:

(?i).+?s/(\d{1,7}).+?>(?:\d+ :|)(\d{2}|)(?:x|)(\d{2}|)(?: - |)(.+?)</a> .(\D{3}/\d{2}/\d{4})..+?k.>(.+?)</font>

But then comes the tricky bit.

Cast members are formatted as, e.g.

<a href='/person/id-1046/Xander+Berkeley'>Xander Berkeley</a> as George Mason

Adding

.+?(<a href=.+?</td>)

to the end of the RegEx simply extracts one long string, but is there a more elegant solution whereby I can extract <name> as <name>, <name> as <name>, ...?

Could someone give me a pointer as where to go on this (if it is feasible)?

Regards

DG

TVRage.txt

[Xenobiologist]

Hi,

did you try to delete all html tags and the split the texts or just give _IEBodyReadText a try.

Mega

[SmOke_N]

Local $s_text = FileRead(@DesktopDir & "\TVRage.txt")
Local $s_pattern = "(?i)(.*?href='/person/id-\d+/.+?'>)([\w\s]+)((?:\s*\(\d+\))*</a>)([\w\s]+)(?:\z|,|</td>.+?\z)"
$s_text = StringTrimRight(StringRegExpReplace($s_text,  $s_pattern, "\2\4, "), 2)
MsgBox(0, 0, $s_text)

Edited August 20, 2008 by SmOke_N

[DobraGolonka]

@Xenobiologist

I didn't mention that I'm getting the data using _INetGetSource - sorry about that.

Local $s_text = FileRead(@DesktopDir & "\TVRage.txt")
Local $s_pattern = "(?i)(.*?href='/person/id-\d+/.+?'>)([\w\s]+)((?:\s*\(\d+\))*</a>)([\w\s]+)(?:\z|,|</td>.+?\z)"
$s_text = StringTrimRight(StringRegExpReplace($s_text,  $s_pattern, "\2\4, "), 2)
MsgBox(0, 0, $s_text)

Wow! That certainly does the trick!

Would you be so kind as to confirm what I *think* is happening?

• (?i) Case-insensitivity flag
• [0] (.*?href='/person/id-\d+/.+?'>) Match the 'href.../nnn/...>' string
• [1] ([\w\s]+) Match any "word" OR whitespace character AND repeat the previous set 1 or more times (Actor Name) until...
• [2] ((?:\s*\(\d+\))*</a>) Non-capturing group: whitespace characters but match any digit (0-9) repeated 1 or more times terminated by </a>
• [3] ([\w\s]+) Match any "word" OR whitespace character AND repeat the previous set 1 or more times (Character Name) until...
• [4] (?:\z|,|</td>.+?\z) Non-capturing group: match only at end of string OR Comma OR </td> & match only at end of string
• Finally, replace groups 2 & 4 with " "

I'm struggling a bit with number 4. - it appears to matching on the digits contained in the 'href' string??

Kind regards

DG

[SmOke_N]

@Xenobiologist

I didn't mention that I'm getting the data using _INetGetSource - sorry about that.

Wow! That certainly does the trick!

Would you be so kind as to confirm what I *think* is happening?

• (?i) Case-insensitivity flag
• [0] (.*?href='/person/id-\d+/.+?'>) Match the 'href.../nnn/...>' string
• [1] ([\w\s]+) Match any "word" OR whitespace character AND repeat the previous set 1 or more times (Actor Name) until...
• [2] ((?:\s*\(\d+\))*</a>) Non-capturing group: whitespace characters but match any digit (0-9) repeated 1 or more times terminated by </a>
• [3] ([\w\s]+) Match any "word" OR whitespace character AND repeat the previous set 1 or more times (Character Name) until...
• [4] (?:\z|,|</td>.+?\z) Non-capturing group: match only at end of string OR Comma OR </td> & match only at end of string
• Finally, replace groups 2 & 4 with " "

I'm struggling a bit with number 4. - it appears to matching on the digits contained in the 'href' string??

Kind regards

DG

[4] basically is saying to end the matches at the end of the string, a comma, or </td> rest of string

[*] Is close, what I'm telling it is to replace all other matches with nothing, and keep matches 2 and 4 with a space after 4 (the as Name Last Name)