MensBeam
/
lax
3
0
Fork 0
Code Issues 3 Pull Requests Releases Activity
Browse Source

Feed modification dates

master
J. King 4 years ago
parent
c2c7452764
commit
1e46b5ac88
7 changed files with 162 additions and 18 deletions
Whitespace
Show all changes Ignore whitespace when comparing lines Ignore changes in amount of whitespace Ignore changes in whitespace at EOL
Split View
Diff Options
Show Stats Download Patch File Download Diff File
  1. 1
      lib/Parser/JSON/Feed.php
  2. 32
      lib/Parser/XML/Construct.php
  3. 21
      lib/Parser/XML/Feed.php
  4. 2
      tests/cases/AbstractParserTestCase.php
  5. 42
      tests/cases/XML/feed-atom.yaml
  6. 42
      tests/cases/XML/feed-rss1.yaml
  7. 40
      tests/cases/XML/feed-rss2.yaml

1
lib/Parser/JSON/Feed.php
View File

@ -122,6 +122,7 @@ class Feed implements \MensBeam\Lax\Parser\Feed {
* JSON Feed does not have categories at the feed level, so this always returns and empty collection
*/
public function getCategories(): CategoryCollection {
// TODO: the cast extension does add (iTunes) categories to feeds
return new CategoryCollection;
}

32
lib/Parser/XML/Construct.php
View File

@ -19,6 +19,11 @@ abstract class Construct {
protected const TEXT_PLAIN = "plain";
protected const TEXT_HTML = "html";
protected const DATE_ANY = 0;
protected const DATE_LATEST = 1;
protected const DATE_EARLIEST = 2;
protected const DATE_ALL = 3;
/** @var \DOMDocument */
protected $document;
/** @var \DOMXPath */
@ -68,22 +73,39 @@ abstract class Construct {
* Returns null if no suitable nodes were found
*
* @param string $query The XPath query of the nodes to return
* @param bool|null $multi Whether to return multiple results as an array (true) or one result as a date object (false, default)
* @param bool|null $mode Whether to return the first valid date found (DATE_ANY), the earliest chronologically (DATE_EARLIEST), latest chronologically (DATE_LATEST), or all valid dates (DAATE_ALL) in a sorted array
* @param \DOMNode $context The context node for the XPath query
* @return \MensBeam\Lax\Date|array|null
*/
protected function fetchDate(string $query, ?bool $multi = null, \DOMNode $context = null) {
protected function fetchDate(string $query, int $mode = self::DATE_ANY, \DOMNode $context = null) {
$out = [];
$tz = new \DateTimeZone("UTC");
assert(in_array($mode, [self::DATE_ANY, self::DATE_ALL, self::DATE_EARLIEST, self::DATE_LATEST]));
foreach((array) $this->fetchString($query, null, true, $context) as $d) {
if ($d = $this->parseDate($d ?? "")) {
if (!$multi) {
if ($mode === self::DATE_ANY) {
return $d;
} else {
$out[] = $d;
// add the date to the output only if it is a unique moment in time so far
$ts = $d->setTimezone($tz)->format("Y-m-d\TH:i:s.u\Z");
if (!isset($out[$ts])) {
$out[$ts] = $d;
}
}
}
}
return $out ?: null;
// sort the dates earliest to latest and produce an indexed array
ksort($out);
$out = array_values($out);
// return based on requested mode
switch ($mode) {
case self::DATE_ALL:
return $out;
case self::DATE_EARLIEST:
return $out ? $out[0] : null;
case self::DATE_LATEST:
return $out ? array_pop($out) : null;
}
}
/** Returns the first valid URL matching an XPath query. Relative URLs are resolved when possible

21
lib/Parser/XML/Feed.php
View File

@ -88,7 +88,7 @@ class Feed extends Construct implements \MensBeam\Lax\Parser\Feed {
$feed->link = $this->getLink();
$feed->title = $this->getTitle();
$feed->summary = $this->getSummary();
//$feed->dateModified = $this->getDateModified();
$feed->dateModified = $this->getDateModified();
//$feed->icon = $this->getIcon();
//$feed->image = $this->getImage();
//$feed->people = $this->getPeople();
@ -135,6 +135,14 @@ class Feed extends Construct implements \MensBeam\Lax\Parser\Feed {
return $this->getSummaryAtom() ?? $this->getSummaryDC() ?? $this->getSummaryRss1() ?? $this->getSummaryRss2() ?? $this->getSummaryPod();
}
public function getDateModified(): ?Date {
/* fetching a date works differently from other data as only Atom has
well-defined semantics here. Thus the semantics of all the other
formats are equal, and we want the latest date, whatever it is.
*/
return $this->fetchDate("atom:updated", self::DATE_LATEST) ?? $this->fetchDate("dc:date|pubDate|lastBuildDate", self::DATE_LATEST);
}
public function getCategories(): CategoryCollection {
return $this->getCategoriesAtom() ?? $this->getCategoriesRss2() ?? $this->getCategoriesDC() ?? $this->getCategoriesPod() ?? new CategoryCollection;
}
@ -147,10 +155,6 @@ class Feed extends Construct implements \MensBeam\Lax\Parser\Feed {
return $authors->merge($contributors, $editors, $webmasters);
}
public function getDateModified(): ?Date {
return $this->getDateModifiedAtom() ?? $this->getDateModifiedDC() ?? $this->getDateModifiedRss2();
}
public function getEntries(FeedStruct $feed = null): array {
return $this->getEntriesAtom() ?? $this->getEntriesRss1() ?? $this->getEntriesRss2() ?? [];
}
@ -168,6 +172,7 @@ class Feed extends Construct implements \MensBeam\Lax\Parser\Feed {
return $this->fetchString("apple:complete", "(?-i:Yes)") ? true : null; // case-sensitive pattern
}
/** Fetches the "time-to-live" value (a number of minutes before the feed should be re-fetched) from an RSS 2.0 feed */
protected function getSchedIntervalRss2(): ?\DateInterval {
$ttl = (int) $this->fetchString("ttl", "\d+");
if ($ttl) {
@ -176,6 +181,12 @@ class Feed extends Construct implements \MensBeam\Lax\Parser\Feed {
return null;
}
/** Fetches the schedule interval from an RSS feed; this is necessarily approximate:
*
* The interval is defined in the syndication RSS extension as fractions of a period, but PHP only supports integer intervals, so we perform integer divison on the nearest subdivision of a period, returning at least one.
*
* For example, "four times monthly" first assumes a month is 30 days, and divides this by four to yield seven days.
*/
protected function getSchedIntervalRss1(): ?\DateInterval {
$period = $this->fetchString("sched:updatePeriod", "(?:year|month|week|dai|hour)ly");
if ($period) {

2
tests/cases/AbstractParserTestCase.php
View File

@ -76,6 +76,8 @@ class AbstractParserTestCase extends \PHPUnit\Framework\TestCase {
foreach ($output as $k => $v) {
if (in_array($k, ["title", "summary"])) {
$f->$k = $this->makeText($v);
} elseif ($k === "dateModified") {
$f->$k = new Date($v, new \DateTimeZone("UTC"));
} elseif ($k === "people") {
$c = new PersonCollection;
foreach ($v as $m) {

42
tests/cases/XML/feed-atom.yaml
View File

@ -353,3 +353,45 @@ Ignored text constructs:
output:
format: atom
version: '1.0'
Feed date:
input: >
<feed xmlns="http://www.w3.org/2005/Atom">
<updated>2020-03-03T00:00:00Z</updated>
</feed>
output:
format: atom
version: '1.0'
dateModified: '2020-03-03T00:00:00Z'
Multiple feed dates:
input: >
<feed xmlns="http://www.w3.org/2005/Atom">
<updated>2020-03-03T00:00:00Z</updated>
<updated>2020-03-04T00:00:00Z</updated>
</feed>
output:
format: atom
version: '1.0'
dateModified: '2020-03-04T00:00:00Z'
Multiple feed date timezones:
input: >
<feed xmlns="http://www.w3.org/2005/Atom">
<updated>2020-03-03T00:00:00Z</updated>
<updated>2020-03-03T00:00:00-04:00</updated>
<updated>2020-03-03T01:00:00Z</updated>
</feed>
output:
format: atom
version: '1.0'
dateModified: '2020-03-03T00:00:00-04:00'
Bogus feed date:
input: >
<feed xmlns="http://www.w3.org/2005/Atom">
<created>2020-03-03T00:00:00Z</created>
</feed>
output:
format: atom
version: '1.0'

42
tests/cases/XML/feed-rss1.yaml
View File

@ -194,9 +194,9 @@ Feed link:
input: >
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://purl.org/rss/1.0/">
<channel>
<link/>
<link>http://[example.net]/</link>
<link>http://example.com/</link>
<link/>
<link>http://[example.net]/</link>
<link>http://example.com/</link>
</channel>
</rdf:RDF>
output:
@ -208,7 +208,7 @@ Feed title 1:
input: >
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://purl.org/rss/1.0/">
<channel>
<title> Loose text</title>
<title> Loose text</title>
</channel>
</rdf:RDF>
output:
@ -221,7 +221,7 @@ Feed title 2:
input: >
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://purl.org/rss/1.0/">
<channel>
<title xml:base="https://example.com/"> Loose text</title>
<title xml:base="https://example.com/"> Loose text</title>
</channel>
</rdf:RDF>
output:
@ -235,7 +235,7 @@ DC title:
input: >
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://purl.org/rss/1.0/" xmlns:dc="http://purl.org/dc/elements/1.1/">
<channel>
<dc:title xml:base="https://example.com/"> Plain text</dc:title>
<dc:title xml:base="https://example.com/"> Plain text</dc:title>
</channel>
</rdf:RDF>
output:
@ -248,7 +248,7 @@ Feed summary:
input: >
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://purl.org/rss/1.0/">
<channel>
<description>Loose text</description>
<description>Loose text</description>
</channel>
</rdf:RDF>
output:
@ -261,7 +261,7 @@ DC summary:
input: >
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://purl.org/rss/1.0/" xmlns:dc="http://purl.org/dc/elements/1.1/">
<channel>
<dc:description>Plain text</dc:description>
<dc:description>Plain text</dc:description>
</channel>
</rdf:RDF>
output:
@ -269,3 +269,29 @@ DC summary:
version: '1.0'
summary:
plain: 'Plain text'
DC date:
input: >
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://purl.org/rss/1.0/" xmlns:dc="http://purl.org/dc/elements/1.1/">
<channel>
<dc:date>2020-03-03T00:00:00Z</dc:date>
</channel>
</rdf:RDF>
output:
format: rdf
version: '1.0'
dateModified: '2020-03-03T00:00:00Z'
Multiple DC dates:
input: >
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns="http://purl.org/rss/1.0/" xmlns:dc="http://purl.org/dc/elements/1.1/">
<channel>
<dc:date>2020-03-03T00:00:00Z</dc:date>
<dc:date>2020-03-03T00:00:00-04:00</dc:date>
<dc:date>2020-03-03T01:00:00Z</dc:date>
</channel>
</rdf:RDF>
output:
format: rdf
version: '1.0'
dateModified: '2020-03-03T00:00:00-04:00'

40
tests/cases/XML/feed-rss2.yaml
View File

@ -330,3 +330,43 @@ Feed summary:
format: rss
summary:
loose: 'Loose text'
Feed publication date:
input: >
<rss><channel>
<pubDate>2020-03-03T00:00:00Z</pubDate>
</channel></rss>
output:
format: rss
dateModified: '2020-03-03T00:00:00Z'
Feed build date:
input: >
<rss><channel>
<lastBuildDate>2020-03-03T00:00:00Z</lastBuildDate>
</channel></rss>
output:
format: rss
dateModified: '2020-03-03T00:00:00Z'
Multiple dates 1:
input: >
<rss><channel>
<pubDate>2020-03-03T00:00:00Z</pubDate>
<pubDate>2020-03-03T00:00:00-04:00</pubDate>
<lastBuildDate>2020-03-03T00:00:00Z</lastBuildDate>
</channel></rss>
output:
format: rss
dateModified: '2020-03-03T00:00:00-04:00'
Multiple dates 2:
input: >
<rss><channel>
<pubDate>2020-03-03T00:00:00Z</pubDate>
<lastBuildDate>2020-03-03T00:00:00-04:00</lastBuildDate>
<pubDate>2020-03-03T00:00:00Z</pubDate>
</channel></rss>
output:
format: rss
dateModified: '2020-03-03T00:00:00-04:00'

Write Preview
Loading…
Cancel
Save