Embedded Conditions Not Working Correctly

[PHenry1026]

The regex below, with the embedded condition, (?(?= – ) – 0?([1-9]|1[0-2]):0([1-9])\s*?([AP])\.?M\.?), should pick-up the time 4:15 – 5:05 p.m.
in the text below and it does.


(?#dd:To)(?m)(?<=^|\s)['"‘“\p{Pi}\p{Ps}]{0,2}0?([1-9]|1[0-2]):([1-5][0-9])(?(?= – ) – 0?([1-9]|1[0-2]):0([1-9])\s*?([AP])\.?M\.?)(?=\s|$)



$1 $2 to $3 o $4 $5.M



<< start of text >>



Arts & Letters Day

Romance and Mystery Novels: an Intimate Conversation with Annette Blair

In a cozy living room setting Annette Blair discusses the professional life of a writer of romance and mystery novels.

Wednesday, April 6th

4:15 – 5:05 p.m.

Conference Center, Second Floor, Room K-422

<< end of text >>





However if encounters the following text it will also incorrectly affect and mispronounce the time, 8:19 PM PDT
.



<< start of 2nd text >>



Yahoo! Breaking News Friday, April 8, 2011, 8:19 PM PDT

WASHINGTON (AP) Senate passes short-term spending bill to keep the government open



<< end of 2nd text >>




8:19 PM PDT should be unaffected by the above regex since the lookahead condition, (?= – ), should have failed. I have been able to get it to work properly by introducing an else condition as follows:



(?#dd:To)(?m)(?<=^|\s)['"‘“\p{Pi}\p{Ps}]{0,2}0?([1-9]|1[0-2]):([1-5][0-9])(?(?= – ) – 0?([1-9]|1[0-2]):0([1-9])\s*?([AP])\.?M\.?|he is a Jackass)(?=\s|$)



$1 $2 to $3 o $4 $5.M


However, it seems that since the condition is not met the regex should have failed without the else condition.


Percy Henry

[PHenry1026]

I have given some more thought to this problem and on further reflection, it seems that the regex engine is working correctly. Without, the else clause the embedded condition, (?(?= – ) – 0?([1-9]|1[0-2]):0([1-9])\s*?([AP])\.?M\.?), is optional and the non-optional part of the regex, ([1-9]|1[0-2]):([1-5][0-9]), correctly matches 8:19 in the second text sample; although this was not intended.

I guess the lesson of this exercise, for me, is that if you are going to use embedded conditions in TA3, you need to make up a dummy else statement that has a low probability of matching.

Percy Henry