Crispi
/
dogstats
Archived
2
0
Fork 0
Code Issues 1 Pull Requests Projects Wiki Activity

crawler update
Build and push / Pulling repo on server (push) Successful in 3s Details

This commit is contained in:
Chris 2023-12-11 19:29:14 +00:00
parent 4fc04c2793
commit b00f652283
5 changed files with 13 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
crawler/crawler.php
View File

@ -1,7 +1,10 @@
<?php
//$ergebnisse = file_get_contents('https://www.dognow.at/ergebnisse/');
//file_put_contents('tmp/ergebnisse.html', $ergebnisse);
if(!file_exists('tmp/ergebnisse.html'))
{
$ergebnisse = file_get_contents('https://www.dognow.at/ergebnisse/');
file_put_contents('tmp/ergebnisse.html', $ergebnisse);
}
$GLOBALS['db'] = new SQLite3('data.db');
if(!$GLOBALS['db']) exit("Error loading database");
@ -177,6 +180,7 @@ function getResults($run,$event)
function analyzeResultCSV($csvfile,$run,$event)
{
if(!file_exists($csvfile)) die(" ERR: File $csvfile not found");
$csv = array_map('str_getcsv', file($csvfile));
//prepare header for database
@ -229,8 +233,10 @@ function convertPDFtoCSV($pdf,$targetname)
}
function analyze($pdf) {
echo " [i] Analyzing $pdf\n";
$cmd = "java -jar tabula-1.0.5-jar-with-dependencies.jar -f CSV $pdf";
$output = shell_exec($cmd);
//var_dump($output);
return $output;
}

BIN
crawler/data.db
View File

Binary file not shown.

2
crawler/tmp/.gitignore vendored
View File

@ -1,2 +1,2 @@
*
*.html
!.gitignore

2
crawler/tmp/csv/.gitignore vendored Normal file
View File

@ -0,0 +1,2 @@
*
!.gitignore

2
crawler/tmp/results/.gitignore vendored Normal file
View File

@ -0,0 +1,2 @@
*
!.gitignore