Last active
July 20, 2018 11:53
-
-
Save nattaylor/8ed8c65eda4dc7c966498cf70f1008fd to your computer and use it in GitHub Desktop.
Zoning Board of Appeal Decisions - City of Boston
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
<!DOCTYPE html> | |
<html> | |
<head> | |
<title>Zoning Board of Appeal Decisions - City of Boston</title> | |
<meta name="viewport" content="width=device-width, initial-scale=1"> | |
<style> | |
body { | |
font-family: sans-serif; | |
} | |
table { | |
border-collapse: collapse; | |
} | |
td, th { | |
border:1px solid #ccc; | |
} | |
</style> | |
</head> | |
<body> | |
<h1>Zoning Board of Appeal Decisions - City of Boston</h1> | |
<p>This an aggregation of the information available at <a href="https://www.boston.gov/departments/inspectional-services/zoning-board-appeal-decisions">https://www.boston.gov/departments/inspectional-services/zoning-board-appeal-decisions</a>. Appeals that haven't been fully aggregated yet may appear in the tables not in the details, and the details may contain typos due to the aggregation process. For more information, contact <a href="mailto:nattaylor@gmail.com">nattaylor@gmail.com</a>.</p> | |
<?php | |
$appeals = array_map(function($line){return explode("\t",$line);} ,explode("\n", file_get_contents('appeals.tsv'))); | |
$totals = array(); | |
foreach($appeals as $appeal) { | |
if($appeal[2] == "defer") continue; | |
if(!isset($totals[$appeal[4]])) { | |
$totals[$appeal[4]] = array("approvals"=>0, "denials"=>0, "total"=>0, "ratio"=>0.0); | |
} | |
if($appeal[2]=="approve") { | |
$totals[$appeal[4]]["approvals"]+=1; | |
} else if ($appeal[2]=="deny") { | |
$totals[$appeal[4]]["denials"]+=1; | |
} | |
$totals[$appeal[4]]["total"]+=1; | |
$totals[$appeal[4]]["ratio"]=strval(round($totals[$appeal[4]]["approvals"]/$totals[$appeal[4]]["total"],2)*100)."%"; | |
} | |
ksort($totals); | |
echo "<p><strong>Table of Contents</strong></p><ul>\n"; | |
foreach(range(1,21) as $ward) { | |
echo "<li><a href=\"#ward$ward\">Ward $ward</a></li>"; | |
} | |
echo "<li><a href=\"#details\">Details</a></li>"; | |
echo "</ul>\n"; | |
echo "<h2>Summary</h2>"; | |
echo "<table>\n<thead><tr><th>Ward</th><th>Approvals</th><th>Denials</th><th>Total</th><th>Rate</th></tr></thead>\n<tbody>\n"; | |
foreach($totals as $ward=>$total) { | |
echo "<tr><td>$ward</td><td>".implode("</td><td>", $total)."</td></tr>".PHP_EOL; | |
} | |
echo "</tbody></table>".PHP_EOL; | |
foreach(range(1,21) as $ward) { | |
echo "<h2 id=\"ward$ward\">Ward $ward</h2>".PHP_EOL; | |
echo "<table>\n<thead><tr><th>Appeal</th><th>Date</th><th>Vote</th><th>Address</th></tr></thead>\n<tbody>\n"; | |
foreach($appeals as $appeal) { | |
if($appeal[2] == "defer") continue; | |
if($appeal[4] == "Ward $ward") { | |
array_pop($appeal); | |
$appeal[0] = "<a href=\"#{$appeal[0]}\">{$appeal[0]}</a>"; | |
echo "<tr><td>".implode("</td><td>", $appeal)."</td></tr>".PHP_EOL; | |
} | |
} | |
echo "</tbody></table>".PHP_EOL; | |
} | |
echo "<h2 id=\"details\">Details</h2>\n"; | |
exec("php -f process.php list.txt",$output); | |
echo implode("\n", $output); | |
?> | |
<a href="#">Return to top ⇑</a> | |
</body> | |
</html> |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
<?php | |
/** Parses the data from https://www.boston.gov/departments/inspectional-services/zoning-board-appeal-decisions | |
* $str needs to be of this form: | |
January 30, 2018 | |
APPROVED DECISIONS | |
BOA-775971 268 Bremen Street, Ward 1 | |
BOA-766613 109 Beech Street, Ward 20 | |
BOA-734095 34 Concord Square, Ward 4 | |
DENIED DECISIONS | |
BOA-747215 117 Bolton Street, Ward 6 | |
BOA-700987 5 Cypher Street, Ward 6 | |
BOA-736227 18 Greenwich Street, Ward 15 | |
*/ | |
$str = ""; | |
//Split on date | |
$hearings = preg_split("/\n([A-Z][a-z]* [0-9]{1,2}, [0-9]{4})\n/", $str, -1, PREG_SPLIT_DELIM_CAPTURE); | |
for($i=1; $i<count($hearings); $i+=2) { | |
$monthMapping = array("January"=>'01', "February"=>'02', "March"=>'03', "April"=>'04', "May"=>'05', "June"=>'06', "July"=>'07', "August"=>'08', "September"=>'09', "October"=>'10', "November"=>'11', "December"=>'12'); | |
preg_match("/(January|February|March|April|May|June|July|August|September|October|November|December) ([0-9]{1,2}), ([0-9]{4})/", $hearings[$i], $dateMatches); | |
$date = "{$dateMatches[3]}-{$monthMapping[$dateMatches[1]]}-{$dateMatches[2]}"; | |
$splitHearings = preg_split("/DENIED DECISIONS/", $hearings[$i+1]); | |
$splitHearings[0] = str_replace("APPROVED DECISIONS\n", "", $splitHearings[0]); | |
for($j=0;$j<2;$j++) { | |
if(!isset($splitHearings[$j])) {continue;} | |
$appeals = explode("\n", trim($splitHearings[$j])); | |
$vote = ($j == 0) ? "approve" : "deny"; | |
foreach($appeals as $appeal) { | |
$parts = preg_split("/(BOA-[0-9]*) /", $appeal, -1, PREG_SPLIT_DELIM_CAPTURE); | |
//$parts = array_slice($parts, 1); | |
$parts = array_slice(array_merge(array($parts[1], $date, $vote), preg_split("/, (Ward.*)/", $parts[2], -1, PREG_SPLIT_DELIM_CAPTURE)),0,-1); | |
echo "<tr><td>".implode("</td><td>", $parts)."</td></tr>".PHP_EOL; | |
} | |
} | |
} |
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
<?php | |
/** | |
* Parse a Zoning Board hearing minutes into structured data. | |
* - https://www.boston.gov/departments/inspectional-services/zoning-board-appeal#meeting-minutes | |
* | |
*/ | |
$files = $argv[1] == "list.txt" ? explode("\n", file_get_contents("list.txt")) : array($argv[1]); | |
$mergedCases = array(); | |
foreach ($files as $file) { | |
/** | |
* Ugly date parsing. | |
*/ | |
$monthMapping = array("january"=>'01', "february"=>'02', "march"=>'03', "april"=>'04', "may"=>'05', "june"=>'06', "july"=>'07', "august"=>'08', "september"=>'09', "october"=>'10', "november"=>'11', "december"=>'12'); | |
preg_match("/(january|february|march|april|may|june|july|august|september|october|november|december)_([0-9]{1,2})_([0-9]{4})/", $file, $dateMatches); | |
$date = "{$dateMatches[3]}-{$monthMapping[$dateMatches[1]]}-{$dateMatches[2]}"; | |
preg_match_all("/.*?\nCase:.*?/s", preg_replace("/\n[0-9]{1,2}\n\n/", " ", strip_tags(file_get_contents($file))), $output_array); | |
/** | |
* Loop to insert newlines as delimiters | |
*/ | |
$split = array_map(function($case){ | |
$case = str_replace("\n", " ", $case); | |
$fields = array( | |
"Case", | |
"Address", | |
"Article(s)", | |
"Purpose", | |
"Discussion", | |
"Testimony", | |
"Documents/Exhibits", | |
"Discussion/Vote", | |
"Vote", | |
"Applicant", | |
"Ward" | |
); | |
foreach($fields as $field) { | |
$case = str_replace($field.": ", "\n$field: ", $case); | |
} | |
//Special Cases | |
$case = str_replace("Address:", "\nAddress: ", $case); | |
$case = str_replace("Votes:", "\nVote: ", $case); | |
$case = preg_replace("/, Ward ([0-9]{1,2})/", "\nWard: $1", $case); | |
$case = preg_replace("/Case:$/", "", $case); | |
return "Case:$case"; | |
}, $output_array[0]); | |
/** | |
* Loop to parse out the values | |
*/ | |
$cases = array_map(function($case) use ($date) { | |
$parsedCase = array(); | |
$lines = explode("\n", $case); | |
$fields = array( | |
"Case", | |
"Address", | |
"Article\(s\)", | |
"Purpose", | |
"Discussion", | |
"Testimony", | |
"Documents\/Exhibits", | |
"Discussion\/Vote", | |
"Vote", | |
"Applicant", | |
"Ward" | |
); | |
foreach($lines as $line) { | |
foreach($fields as $field) { | |
preg_match("/^$field: (.*)/", $line, $matches); | |
if( count($matches)>0 ) { | |
$parsedCase[$field]=rtrim($matches[1]," ,"); | |
if($field == "Vote") { | |
if (strstr($matches[1], "deny") || strstr($matches[1], "denial")) { | |
$parsedCase['vote_heuristic']='deny'; | |
} else if (strstr($matches[1], "defer")) { | |
$parsedCase['vote_heuristic']='defer'; | |
} else if (strstr($matches[1], "approv")) { | |
$parsedCase['vote_heuristic']='approve'; | |
}else { | |
$parsedCase['vote_heuristic']='unknown'; | |
} | |
} | |
continue; | |
} | |
} | |
} | |
$parsedCase['Date']=$date; | |
return $parsedCase; | |
}, $split); | |
if(count($files)==1) { | |
foreach ($cases as $case) { | |
if ($case['vote_heuristic'] == 'defer') { | |
continue; | |
} | |
//case, date, vote_heuristic, address, ward | |
$structuredCase = array($case['Case'],$case['Date'],$case['vote_heuristic'],$case['Address'],"Ward ".$case['Ward']); | |
echo implode(" ", $structuredCase).PHP_EOL; | |
} | |
exit(); | |
} else { | |
$mergedCases = array_merge($cases, $mergedCases); | |
} | |
} | |
foreach ($mergedCases as $case) { | |
$fields = array( | |
"Date", | |
"Ward", | |
"Applicant", | |
"Address", | |
"Article\(s\)", | |
"Purpose", | |
"Discussion", | |
"Testimony", | |
"Documents\/Exhibits", | |
"Discussion\/Vote", | |
"Vote" | |
); | |
if(strlen($case["Case"])>10) continue; | |
echo "<h3 id=\"{$case["Case"]}\">{$case["Case"]}</h3>\n"; | |
foreach ($fields as $field) { | |
if(isset($case[$field])) { | |
echo "<p><strong>$field:</strong> {$case[$field]}</p>\n"; | |
} | |
} | |
} |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment