Prepare 0.3.0 Release

bakame-php · Sep 28, 2023 · 1506a96 · 1506a96
1 parent e415048
commit 1506a96
Show file tree

Hide file tree

Showing 5 changed files with 89 additions and 25 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -2,7 +2,7 @@
 
 All Notable changes to `bakame/html-table` will be documented in this file.
 
-## [0.3.0](https://github.com/bakame-php/html-table/compare/0.2.0...0.3.0) - 2023-09-27
+## [0.3.0](https://github.com/bakame-php/html-table/compare/0.2.0...0.3.0) - 2023-09-28
 
 ### Added
 
@@ -14,6 +14,7 @@ All Notable changes to `bakame/html-table` will be documented in this file.
 
 - Improve identifier validation for `Parser::tablePosition`
 - Remove the `$tableOffset` property.
+- `tableHeader` can now re-arrange the table column and remove any unwanted column.
 
 ### Deprecated
 

diff --git a/README.md b/README.md
@@ -228,9 +228,19 @@ $parser = Parser::new()->tableHeader(['rank', 'team', 'winner']);
 ```
 
 **If you specify a non-empty array as the table header, it will take precedence over any other table header related options.**
-
 **Because it is a tabular data each cell MUST be unique otherwise an exception will be thrown**
 
+You can skip some of the source column by skipping them by their offsets
+
+```php
+use Bakame\HtmlTable\Parser;
+use Bakame\HtmlTable\Section;
+
+$parser = Parser::new()->tableHeader([3 => 'rank',  7 => 'winner', 5 => 'team']);
+// only 3 column will be extracted the 4th, 6th and 8th column
+// and re-arrange as 'rank' first and 'team' last
+```
+
 ### includeSection and excludeSection
 
 Tells which section should be parsed based on the `Section` enum
@@ -239,8 +249,8 @@ Tells which section should be parsed based on the `Section` enum
 use Bakame\HtmlTable\Parser;
 use Bakame\HtmlTable\Section;
 
-$parser = Parser::new()->includeSection(Section::thead); // thead is included during parsing
-$parser = Parser::new()->excludeSection(Section::tr);    // table direct tr children are not included during parsing
+$parser = Parser::new()->includeSection(Section::tbody);  // thead and tfoot are included during parsing
+$parser = Parser::new()->excludeSection(Section::tr, Section::tfoot); // table direct tr children and tfoot are not included during parsing
 ```
 
 **By default, the `thead` section is not parse. If a `thead` row is selected to be the header, it will

diff --git a/composer.json b/composer.json
@@ -27,20 +27,22 @@
   ],
   "require": {
     "ext-dom": "*",
-    "ext-json": "*",
     "ext-libxml": "*",
     "ext-mbstring": "*",
     "ext-simplexml": "*",
     "league/csv": "^9.11.0"
   },
   "require-dev": {
+    "ext-curl": "*",
     "ext-xdebug": "*",
     "friendsofphp/php-cs-fixer": "^v3.28.0",
+    "laravel/prompts": "^0.1.9",
     "phpstan/phpstan": "^1.10.35",
     "phpstan/phpstan-deprecation-rules": "^1.1.4",
     "phpstan/phpstan-phpunit": "^1.3.14",
     "phpstan/phpstan-strict-rules": "^1.5.1",
     "phpunit/phpunit": "^10.3.5",
+    "symfony/css-selector": "^6.3",
     "symfony/var-dumper": "^6.3.4"
   },
   "autoload": {

diff --git a/src/Parser.php b/src/Parser.php
@@ -14,20 +14,15 @@
 use Iterator;
 use League\Csv\ResultSet;
 use League\Csv\SyntaxError;
-use League\Csv\TabularDataReader;
 use SimpleXMLElement;
 use Stringable;
 
-use function array_combine;
 use function array_fill;
 use function array_filter;
 use function array_key_exists;
 use function array_merge;
-use function array_pad;
 use function array_shift;
-use function array_slice;
 use function array_unique;
-use function count;
 use function fclose;
 use function fopen;
 use function in_array;
@@ -193,10 +188,12 @@ public function tableHeaderPosition(Section $section, int $offset = 0): self
         };
     }
 
-    public function includeSection(Section $section): self
+    public function includeSection(Section ...$sections): self
     {
         $includedSections = $this->includedSections;
-        $includedSections[$section->value] = 1;
+        foreach ($sections as $section) {
+            $includedSections[$section->value] = 1;
+        }
 
         return match ($this->includedSections) {
             $includedSections => $this,
@@ -213,10 +210,12 @@ public function includeSection(Section $section): self
         };
     }
 
-    public function excludeSection(Section $section): self
+    public function excludeSection(Section ...$sections): self
     {
         $includedSections = $this->includedSections;
-        unset($includedSections[$section->value]);
+        foreach ($sections as $section) {
+            unset($includedSections[$section->value]);
+        }
 
         return match ($this->includedSections) {
             $includedSections => $this,
@@ -322,7 +321,7 @@ public function tableCaption(?string $caption = null): self
      * @throws ParserError
      * @throws SyntaxError
      */
-    public function parseFile(mixed $filenameOrStream, $filenameContext = null): TabularDataReader
+    public function parseFile(mixed $filenameOrStream, $filenameContext = null): Table
     {
         if (is_resource($filenameOrStream)) {
             return $this->parseHtml($this->streamToString($filenameOrStream));
@@ -374,7 +373,7 @@ public function parseHtml(DOMDocument|DOMElement|SimpleXMLElement|Stringable|str
         $result = $xpath->query('(//caption)[1]');
         $caption = $result->item(0)?->nodeValue ?? $this->caption;
 
-        return new Table(new ResultSet($this->extractTableContents($xpath, $header), $header), $caption);
+        return new Table(new ResultSet($this->extractTableContents($xpath, $header), array_values($header)), $caption);
     }
 
     /**
@@ -564,16 +563,30 @@ private function extractRecord(DOMElement $tr, array &$rowSpanIndices = []): arr
      */
     private function formatRecord(array $record, array $header): array
     {
-        $cellCount = count($header);
-        $record = match ($cellCount) {
-            0 => $record,
-            count($record) => array_combine($header, $record),
-            default => array_combine($header, array_slice(array_pad($record, $cellCount, ''), 0, $cellCount)),
+        $record = match ([]) {
+            $header => $record,
+            default => $this->combineArray($record, $header),
         };
 
         return match (null) {
             $this->formatter => $record,
             default => ($this->formatter)($record),
         };
     }
+
+    /**
+     * @param array<string> $record
+     * @param array<string> $header
+     *
+     * @return array<string, string|null>
+     */
+    private function combineArray(array $record, array $header): array
+    {
+        $row = [];
+        foreach ($header as $offset => $value) {
+            $row[$value] = $record[$offset] ?? null;
+        }
+
+        return $row;
+    }
 }
diff --git a/src/ParserTest.php b/src/ParserTest.php
@@ -47,9 +47,7 @@ public function it_will_return_the_same_options(): void
             $parser
                 ->tablePosition(0)
                 ->tableHeaderPosition(Section::thead, 0)
-                ->includeSection(Section::tbody)
-                ->includeSection(Section::tfoot)
-                ->includeSection(Section::tr)
+                ->includeSection(Section::tbody, Section::tfoot, Section::tr)
                 ->tableHeader([])
                 ->resolveTableHeader()
                 ->ignoreXmlErrors()
@@ -286,6 +284,47 @@ public function it_will_use_the_submitted_headers(): void
         ], $table->first());
     }
 
+
+    #[Test]
+    public function it_will_rearrange_the_content_with_table_header(): void
+    {
+        $html = <<<TABLE
+<table class="table-csv-data" id="testb">
+<tfoot>
+<tr data-record-offset="5"><td title="prenoms">Abel</td><td title="nombre">14</td><td title="sexe">M</td><td title="annee">2004</td></tr>
+<tr data-record-offset="6"><td title="prenoms">Abiga</td><td title="nombre">6</td><td title="sexe">F</td><td title="annee">2004</td></tr>
+<tr data-record-offset="7"><td title="prenoms">Aboubacar</td><td title="nombre">8</td><td title="sexe">M</td><td title="annee">2004</td></tr>
+<tr data-record-offset="8"><td title="prenoms">Aboubakar</td><td title="nombre">6</td><td title="sexe">M</td><td title="annee">2004</td></tr>
+</tfoot>
+</table>
+TABLE;
+
+        $header = [3 => 'Annee', 2 => 'Sexe', 0 => 'Firstname', 1 => 'Count'];
+        $table = Parser::new()
+            ->tableHeader($header)
+            ->parseHtml($html);
+
+        self::assertSame($table->getHeader(), array_values($header));
+        self::assertSame([
+            'Annee' => '2004',
+            'Sexe' => 'M',
+            'Firstname' => 'Abel',
+            'Count' => '14',
+        ], $table->first());
+
+        $header = [3 => 'Annee', 0 => 'Firstname', 1 => 'Count'];
+        $table = Parser::new()
+            ->tableHeader($header)
+            ->parseHtml($html);
+
+        self::assertSame($table->getHeader(), array_values($header));
+        self::assertSame([
+            'Annee' => '2004',
+            'Firstname' => 'Abel',
+            'Count' => '14',
+        ], $table->first());
+    }
+
     #[Test]
     public function it_will_duplicate_colspan_data(): void
     {
@@ -391,7 +430,6 @@ public function it_will_found_no_header_in_any_section(): void
         self::assertSame([], $table->getHeader());
     }
 
-
     #[Test]
     public function it_will_use_the_table_footer(): void
     {