From 186c8895131da068621c8ef1bb51585b1283a359 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?=E2=80=9CNaeel=E2=80=9D?= <tazet@narod.ru>
Date: Thu, 18 Jun 2026 16:15:00 +0400
Subject: [PATCH] =?UTF-8?q?v1.0.5:=20parser.cfm=20=E2=80=94=20PDF=20=D1=87?=
 =?UTF-8?q?=D0=B5=D1=80=D0=B5=D0=B7=20PDFBox,=20TXT?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 index.cfm  |   2 +-
 parser.cfm | 116 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 117 insertions(+), 1 deletion(-)
 create mode 100644 parser.cfm

diff --git a/index.cfm b/index.cfm
index 3cc8b01..ef864a8 100644
--- a/index.cfm
+++ b/index.cfm
@@ -1 +1 @@
-<cfsetting showdebugoutput=no enablecfoutputonly=yes><cfoutput>OK v1.0.4</cfoutput>
+<cfsetting showdebugoutput=no enablecfoutputonly=yes><cfoutput>OK v1.0.5</cfoutput>
diff --git a/parser.cfm b/parser.cfm
new file mode 100644
index 0000000..13d4117
--- /dev/null
+++ b/parser.cfm
@@ -0,0 +1,116 @@
+<cfsetting showdebugoutput="no" enablecfoutputonly="true">
+<cfheader name="Content-Type" value="application/json; charset=utf-8">
+
+<cfparam name="url.doc_id" default="">
+
+<cfset result = {ok: false, error: ""}>
+
+<cftry>
+    <cfif NOT len(url.doc_id)>
+        <cfset result.error = "doc_id required">
+    <cfelse>
+        <!--- Читаем документ из БД --->
+        <cfquery name="doc" datasource="baza">
+            SELECT id, filename, mime_type, original_bytes, status
+            FROM documents
+            WHERE id = <cfqueryparam value="#url.doc_id#" cfsqltype="cf_sql_varchar">
+        </cfquery>
+
+        <cfif doc.recordCount EQ 0>
+            <cfset result.error = "document not found">
+        <cfelse>
+            <cfset elements = []>
+            <cfset errors = []>
+
+            <!--- PDF --->
+            <cfif doc.mime_type EQ "application/pdf">
+                <cftry>
+                    <!--- Base64 → байты --->
+                    <cfset fileBytes = binaryDecode(doc.original_bytes, "base64")>
+                    <!--- Байты → InputStream --->
+                    <cfset bais = createObject("java", "java.io.ByteArrayInputStream").init(fileBytes)>
+                    <!--- Открываем PDF --->
+                    <cfset pdfDoc = createObject("java", "org.apache.pdfbox.pdmodel.PDDocument").load(bais)>
+                    <cfset stripper = createObject("java", "org.apache.pdfbox.text.PDFTextStripper")>
+
+                    <cfset pageCount = pdfDoc.getNumberOfPages()>
+                    <cfloop from="1" to="#pageCount#" index="pn">
+                        <cfset stripper.setStartPage(pn)>
+                        <cfset stripper.setEndPage(pn)>
+                        <cfset pageText = stripper.getText(pdfDoc)>
+                        <cfif len(trim(pageText))>
+                            <cfloop list="#pageText#" index="line" delimiters="#chr(10)#">
+                                <cfset line = trim(line)>
+                                <cfif len(line)>
+                                    <cfset arrayAppend(elements, {type: "paragraph", style: "", text: line, page: pn})>
+                                </cfif>
+                            </cfloop>
+                        </cfif>
+                    </cfloop>
+
+                    <cfset pdfDoc.close()>
+                    <cfset bais.close()>
+
+                    <cfcatch>
+                        <cfset arrayAppend(errors, "PDFBox: " & cfcatch.message)>
+                    </cfcatch>
+                </cftry>
+
+            <!--- DOCX/DOC — POI нет, пропускаем --->
+            <cfelseif doc.mime_type EQ "application/vnd.openxmlformats-officedocument.wordprocessingml.document"
+                      OR doc.mime_type EQ "application/msword">
+                <cfset arrayAppend(errors, "DOCX/DOC parsing not available (POI not installed)")>
+
+            <!--- TXT — plain text --->
+            <cfelseif doc.mime_type EQ "text/plain" OR doc.mime_type EQ "application/octet-stream">
+                <cftry>
+                    <cfset fileBytes = binaryDecode(doc.original_bytes, "base64")>
+                    <cfset text = createObject("java", "java.lang.String").init(fileBytes, "UTF-8")>
+                    <cfloop list="#text#" index="line" delimiters="#chr(10)#">
+                        <cfset line = trim(line)>
+                        <cfif len(line)>
+                            <cfset arrayAppend(elements, {type: "paragraph", style: "", text: line, page: 1})>
+                        </cfif>
+                    </cfloop>
+                    <cfcatch>
+                        <cfset arrayAppend(errors, "TXT: " & cfcatch.message)>
+                    </cfcatch>
+                </cftry>
+
+            <cfelse>
+                <cfset arrayAppend(errors, "unsupported mime_type: " & doc.mime_type)>
+            </cfif>
+
+            <!--- Сохраняем результат --->
+            <cfset newStatus = arrayLen(elements) GT 0 ? "parsed" : "error">
+            <cfif arrayLen(errors) GT 0 AND arrayLen(elements) EQ 0>
+                <cfset newStatus = "error">
+            </cfif>
+
+            <cfquery datasource="baza">
+                UPDATE documents
+                SET elements_json = <cfqueryparam value="#serializeJSON(elements)#" cfsqltype="cf_sql_varchar">::jsonb,
+                    status = <cfqueryparam value="#newStatus#" cfsqltype="cf_sql_varchar">,
+                    error_message = <cfqueryparam value="#arrayLen(errors) ? serializeJSON(errors) : ''#" cfsqltype="cf_sql_varchar">
+                WHERE id = <cfqueryparam value="#url.doc_id#" cfsqltype="cf_sql_varchar">
+            </cfquery>
+
+            <cfset result = {
+                ok: true,
+                doc_id: url.doc_id,
+                filename: doc.filename,
+                mime_type: doc.mime_type,
+                status: newStatus,
+                element_count: arrayLen(elements),
+                page_count: isDefined("pageCount") ? pageCount : 0,
+                errors: errors
+            }>
+        </cfif>
+    </cfif>
+
+    <cfcatch>
+        <cfset result = {ok: false, error: cfcatch.message, detail: cfcatch.detail}>
+    </cfcatch>
+</cftry>
+
+<cfoutput>#serializeJSON(result)#</cfoutput>