Initial commit

.gitignore

@@ -0,0 +1,2 @@
+vendor
+composer.lock

.idea/.gitignore

@@ -0,0 +1,8 @@
+# Default ignored files
+/shelf/
+/workspace.xml
+# Editor-based HTTP Client requests
+/httpRequests/
+# Datasource local storage ignored files
+/dataSources/
+/dataSources.local.xml

.idea/modules.xml

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="ProjectModuleManager">
+    <modules>
+      <module fileurl="file://$PROJECT_DIR$/.idea/text-splitter.iml" filepath="$PROJECT_DIR$/.idea/text-splitter.iml" />
+    </modules>
+  </component>
+</project>

.idea/php.xml

@@ -0,0 +1,20 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project version="4">
+  <component name="MessDetectorOptionsConfiguration">
+    <option name="transferred" value="true" />
+  </component>
+  <component name="PHPCSFixerOptionsConfiguration">
+    <option name="transferred" value="true" />
+  </component>
+  <component name="PHPCodeSnifferOptionsConfiguration">
+    <option name="highlightLevel" value="WARNING" />
+    <option name="transferred" value="true" />
+  </component>
+  <component name="PhpProjectSharedConfiguration" php_language_level="8.3" />
+  <component name="PhpStanOptionsConfiguration">
+    <option name="transferred" value="true" />
+  </component>
+  <component name="PsalmOptionsConfiguration">
+    <option name="transferred" value="true" />
+  </component>
+</project>

.idea/text-splitter.iml

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<module type="WEB_MODULE" version="4">
+  <component name="NewModuleRootManager">
+    <content url="file://$MODULE_DIR$" />
+    <orderEntry type="inheritedJdk" />
+    <orderEntry type="sourceFolder" forTests="false" />
+  </component>
+</module>

README.md

@@ -0,0 +1,4 @@
+# 
+
+ref
+https://github.com/kambo-1st/langchain-php

composer.json

@@ -0,0 +1,17 @@
+{
+    "name": "ivampiresp/text-splitter",
+    "description": "text splitter",
+    "type": "library",
+    "autoload": {
+        "psr-4": {
+            "Ivampiresp\\TextSplitter\\": "src/"
+        }
+    },
+    "authors": [
+        {
+            "name": "ivamp",
+            "email": "im@ivampiresp.com"
+        }
+    ],
+    "require": {}
+}

src/Document.php

@@ -0,0 +1,91 @@
+<?php
+
+namespace Ivampiresp\TextSplitter;
+
+use function explode;
+use function strtolower;
+use function array_values;
+use function array_filter;
+use function count;
+
+use const PHP_EOL;
+
+/**
+ * Interface for interacting with a document.
+ */
+class Document
+{
+    public string $pageContent;
+    public string $lookupStr = '';
+    public int $lookupIndex = 0;
+    public array $metadata = [];
+
+    /**
+     * @param string $pageContent
+     * @param string $lookupStr
+     * @param int    $lookupIndex
+     * @param array  $metadata
+     */
+    public function __construct(
+        string $pageContent,
+        string $lookupStr = '',
+        int $lookupIndex = 0,
+        array $metadata = []
+    ) {
+        $this->pageContent = $pageContent;
+        $this->lookupStr = $lookupStr;
+        $this->lookupIndex = $lookupIndex;
+        $this->metadata = $metadata;
+    }
+
+    /**
+     * Paragraphs of the page.
+     *
+     * @return array
+     */
+    public function paragraphs(): array
+    {
+        return explode(PHP_EOL . PHP_EOL, $this->pageContent);
+    }
+
+    /**
+     * Summary of the page (the first paragraph).
+     *
+     * @return string
+     */
+    public function summary(): string
+    {
+        $paragraphs = $this->paragraphs();
+        return $paragraphs[0];
+    }
+
+    /**
+     * Lookup a term in the page, imitating cmd-F functionality.
+     *
+     * @param string $string
+     *
+     * @return string
+     */
+    public function lookup(string $string): string
+    {
+        if (strtolower($string) !== $this->lookupStr) {
+            $this->lookupStr = strtolower($string);
+            $this->lookupIndex = 0;
+        } else {
+            $this->lookupIndex++;
+        }
+
+        $lookups = array_values(array_filter($this->paragraphs(), function ($p) {
+            return str_contains(strtolower($p), $this->lookupStr);
+        }));
+
+        if (empty($lookups)) {
+            return 'No Results';
+        } elseif ($this->lookupIndex >= count($lookups)) {
+            return 'No More Results';
+        } else {
+            $resultPrefix = '(Result ' . ($this->lookupIndex + 1) . '/' . count($lookups) . ')';
+            return $resultPrefix . ' ' . $lookups[$this->lookupIndex];
+        }
+    }
+}

src/RecursiveCharacterTextSplitter.php

@@ -0,0 +1,91 @@
+<?php
+
+namespace Ivampiresp\TextSplitter;
+
+use Exception;
+use function end;
+use function explode;
+use function str_split;
+use function array_merge;
+
+/**
+ * Implementation of splitting text that looks at characters.
+ * Recursively tries to split by different characters to find one
+ * that works.
+ */
+final class RecursiveCharacterTextSplitter extends TextSplitter
+{
+    private array $separators = ["\n\n", "\n", ' ', ''];
+
+    /**
+     * Create a new TextSplitter.
+     *
+     * @param array $options
+     * @throws Exception
+     */
+    public function __construct(array $options = [])
+    {
+        parent::__construct(
+            $options['chunk_size'] ?? 1000,
+            $options['chunk_overlap'] ?? 200
+        );
+
+        $this->separators = $options['separators'] ?? $this->separators;
+    }
+
+    /**
+     * Split incoming text and return chunks.
+     *
+     * @param string $text
+     *
+     * @return array
+     */
+    public function splitText(string $text): array
+    {
+        $finalChunks = [];
+        // Get appropriate separator to use
+        $separator = end($this->separators);
+        foreach ($this->separators as $_s) {
+            if ($_s == '') {
+                $separator = $_s;
+                break;
+            }
+
+            if (str_contains($text, $_s)) {
+                $separator = $_s;
+                break;
+            }
+        }
+
+        // Now that we have the separator, split the text
+        if ($separator) {
+            $splits = explode($separator, $text);
+        } else {
+            $splits = str_split($text);
+        }
+
+        // Now go merging things, recursively splitting longer texts.
+        $_goodSplits = [];
+        foreach ($splits as $s) {
+            if ($this->lengthFunction($s) < $this->chunkSize) {
+                $_goodSplits[] = $s;
+            } else {
+                if ($_goodSplits) {
+                    $mergedText = $this->mergeSplits($_goodSplits, $separator);
+                    $finalChunks = array_merge($finalChunks, $mergedText);
+                    $_goodSplits = [];
+                }
+
+                $otherInfo = $this->splitText($s);
+                $finalChunks = array_merge($finalChunks, $otherInfo);
+            }
+        }
+
+        if ($_goodSplits) {
+            $mergedText = $this->mergeSplits($_goodSplits, $separator);
+            $finalChunks = array_merge($finalChunks, $mergedText);
+        }
+
+        return $finalChunks;
+    }
+}

src/TextSplitter.php

@@ -0,0 +1,165 @@
+<?php
+
+namespace Ivampiresp\TextSplitter;
+
+use Exception;
+
+abstract class TextSplitter
+{
+    protected int $chunkSize;
+    protected int $chunkOverlap;
+
+    /**
+     * Create a new TextSplitter.
+     *
+     * @param int $chunkSize
+     * @param int $chunkOverlap
+     * @throws Exception
+     */
+    public function __construct(int $chunkSize = 1000, int $chunkOverlap = 200)
+    {
+        if ($chunkOverlap > $chunkSize) {
+            throw new Exception(
+                sprintf(
+                    'Got a larger chunk overlap (%d) than chunk size (%d), should be smaller.',
+                    $chunkOverlap,
+                    $chunkSize
+                )
+            );
+        }
+
+        $this->chunkSize = $chunkSize;
+        $this->chunkOverlap = $chunkOverlap;
+    }
+
+    /**
+     * Split text into multiple components.
+     *
+     * @param string $text
+     *
+     * @return array
+     */
+    abstract public function splitText(string $text): array;
+
+    /**
+     * Create documents from a list of texts.
+     *
+     * @param array $texts
+     * @param ?array $metadata
+     *
+     * @return array
+     */
+    public function createDocuments(array $texts, array $metadata = null): array
+    {
+        $metadata = $metadata ?? array_fill(0, count($texts), array());
+        $documents = [];
+        foreach ($texts as $i => $text) {
+            foreach ($this->splitText($text) as $chunk) {
+                $newDoc = new Document(pageContent:$chunk, metadata:$metadata[$i]);
+                $documents[] = $newDoc;
+            }
+        }
+
+        return $documents;
+    }
+
+    /**
+     * Split documents
+     *
+     * @param array $documents
+     *
+     * @return array
+     */
+    public function splitDocuments(array $documents): array
+    {
+        $texts = array_map(function ($doc) {
+            return $doc->pageContent;
+        }, $documents);
+        $metadata = array_map(function ($doc) {
+            return $doc->metadata;
+        }, $documents);
+        return $this->createDocuments($texts, $metadata);
+    }
+
+    private function joinDocs($docs, $separator): ?string
+    {
+        $text = implode($separator, $docs);
+        $text = trim($text);
+        if ($text === '') {
+            return null;
+        } else {
+            return $text;
+        }
+    }
+
+    /**
+     * We now want to combine these smaller pieces into medium size
+     * chunks to send to the LLM.
+     *
+     * @param iterable $splits
+     * @param string   $separator
+     *
+     * @return array
+     */
+    protected function mergeSplits(iterable $splits, string $separator): array
+    {
+        $separatorLen = strlen($separator);
+
+        $docs = [];
+        $currentDoc = [];
+        $total = 0;
+
+        foreach ($splits as $d) {
+            $len = strlen($d);
+            if ($total + $len + (count($currentDoc) > 0 ? $separatorLen : 0) > $this->chunkSize) {
+                if ($total > $this->chunkSize) {
+                    error_log(
+                        sprintf(
+                            'Created a chunk of size %d, which is longer than the specified %d',
+                            $total,
+                            $this->chunkSize
+                        )
+                    );
+                }
+
+                if (count($currentDoc) > 0) {
+                    $doc = $this->joinDocs($currentDoc, $separator);
+                    if ($doc !== null) {
+                        $docs[] = $doc;
+                    }
+
+                    while (
+                        $total > $this->chunkOverlap
+                        || (
+                            $total + $len + (count($currentDoc) > 0 ? $separatorLen : 0) > $this->chunkSize
+                            && $total > 0
+                        )
+                    ) {
+                        $total -= strlen($currentDoc[0]) + (count($currentDoc) > 1 ? $separatorLen : 0);
+                        array_shift($currentDoc);
+                    }
+                }
+            }
+
+            $currentDoc[] = $d;
+            $total += $len + (count($currentDoc) > 1 ? $separatorLen : 0);
+        }
+
+        $doc = $this->joinDocs($currentDoc, $separator);
+        if ($doc !== null) {
+            $docs[] = $doc;
+        }
+
+        return $docs;
+    }
+
+    /**
+     * @param mixed $s
+     *
+     * @return int
+     */
+    protected function lengthFunction(mixed $s): int
+    {
+        return strlen($s);
+    }
+}