parsing using regular expression

[babygodzilla]

i have a text file that has a pattern of something like ID#, Code, Name, Time, Room#. but once in a while not all of these variables come up in a line. so something like:

please ignore underscores as spaces

12345ABCDaiba, Yu0710-1000145
53154BEC______________TBA____315
__TCRKerr, Hum0200-0400_
21678_____Mama, Papa_____________SAT1

please ignore underscores as spaces

something like that. im sure u get the idea. what i'm doing is extracting the variables using regex and putting them in an array, and later in a database.

the question is, is it possible to have just one expression to accommodate for all these possibilities? if not, what is the minimum number of regex i can have?

THANKS

[babygodzilla]

no takers at all???

[m3rajk]

are the ___ representing of spaces?

have you looked at explode???

[babygodzilla]

yes up there i specified
please ignore underscores as spaces

yea i looked at explode but after discussing it with someone else he suggested to me using regex is a better way.

the reason is that explode separates a string by another string separator. this text file uses just spaces as separators. i cant use that because some variables, like the name of a person, use spaces (in this case between last and first name) and id like to put those in one variable not two.

[babygodzilla]

neone else?

[ppg3412]

Perhaps something like this be executed on every single line?
Note that the text file must be split into an array first with explode("\n",$textdata); (if not already done with file()😉.

This should work as long as the number of spaces between the values (ID# [3] Code [5] Name [5] Time [5] Room#) always remain the same, and the values never include a set of 3 spaces or more.

EDIT: Wasn't behaving upon posting; included in text file.

GL 🙂

[tbach2]

if number of spaces is not the same, why not replace each group of spaces with a tab, and then explode it based on tab?

something along the lines of

preg_replace("/[\s]+/","\t",$row)

(you'll have to play with it, this is just off the top of my head)

[babygodzilla]

this code is good! 🙂

what about NULL characters?

how bout the case of

12345ABCDaiba, Yu0710-1000145
53154BEC______________TBA____315
__TCRKerr, Hum___0200-0400

that third line ends after 0400, so its shorter than the first two. is there a way to put something like (.*|'')
any character any number OR NULL

something like that?

[tbach2]

On the line without the id# (ouch!), is there any character serving as a placeholder for the missing field? If there isn't, you're going to have to go back and fix them manually...

Check out the Pattern Syntax Manual

[babygodzilla]

no there isnt 🙁 the line ends after the character zero...

hmm i was afraid of that...

[tbach2]

actually I was referring to the missing field in front of 'TCR'

[babygodzilla]

oh oh yea yea there are spaces

those underscores are spaces 🙂

[tbach2]

right, but is there any way to know that a field has no value, or are the spaces just stuck together? if the spaces are just stuck together, regex won't be much help.

hmm, if the spaces were inserted to make the columns line up, then each column would start at a known location in the string, right?

if that's the case, you can use some string formatting-- the idea being:

use substr() to pick out a range of characters from the string:

$id = trim(substr($row,0,5));
$code = trim(substr($row,8,3));
etc.

[babygodzilla]

thats correct. each field has its own limit of starting and ending character number. what I'm concern about are the NULL values.
like the case of

12345ABCDaiba, Yu0710-1000145
53154BEC______________TBA____315
__TCRKerr, Hum___0200-0400

this last one is shorter than the first two. what if my code consist of

$room = trim(substr($row,20,5));

i dunno if that really starts at character 20, but just hypothetically. so the code is search for room number starting at character 20, but in reality the string ends at say character 13. would that post a problem? do i have to account for NULL values or EOF?

[tbach2]

substr($row." ________",20,5)

(underscore = space, of course)

[babygodzilla]

wait, what does that do again?

[tbach2]

I just tacked on spaces to the end of the row

[babygodzilla]

oh... how does that help if there are NULLs?

12345ABCDaiba, Yu0710-1000145
53154BEC______________TBA____315
__TCRKerr, Hum___0200-0400(NULLNULLNULLNULL...)

[tbach2]

Wuhl, I don't spose it would... I'm familiar with a null in a database, but what defines a null in a text file? ...That's why I basically ignored it :/

[babygodzilla]

hmmm interesting... 🙂

i thought a NULL would be just like '' u know? two quotes nothing in between so like no value