From 3318500477f9730afca8689158e6ff6939175db6 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Joan=20Fabr=C3=A9gat?= <joan@codeinc.co>
Date: Thu, 22 Feb 2024 12:11:35 -0800
Subject: [PATCH] adds the `format` option

---
 README.md | 12 +++++++-----
 main.mjs  | 20 ++++++++++----------
 2 files changed, 17 insertions(+), 15 deletions(-)

diff --git a/README.md b/README.md
index 66ac60a..f04fcdf 100644
--- a/README.md
+++ b/README.md
@@ -19,7 +19,7 @@ Additional parameters can be sent to customize the conversion process:
 * `lastPage`: The last page to extract. Default is the last page of the document.
 * `password`: The password to unlock the PDF. Default is none.
 * `normalizeWhitespace`: If set to `true`, the server normalizes the whitespace in the extracted text. Default is `true`.
-* `raw`: If set to `true`, the server returns the raw text extracted from the PDF as `text/plain`, else the text is returned as `text/json`. Default is `false`.
+* `format`: The output format. Supported values are `text` (the server returns the raw text as `text/plain`) or `json` (the server returns a JSON object as `text/json`). Default is `text`.
 
 The server returns `200` if the conversion was successful and the images are available in the response body. In case of error, the server returns a `400` status code with a JSON object containing the error message (format: `{error: string}`).
 
@@ -36,16 +36,18 @@ Convert a PDF file to text with a JSON response:
 curl -X POST -F "file=@/path/to/file.pdf" http://localhost:3000/convert -o example.json
 ```
 
-Convert a password-protected PDF file to text with a JSON response:
+Convert a PDF file to text:
 ```bash
-curl -X POST -F "file=@/path/to/file.pdf" -F "password=XXX" http://localhost:3000/convert -o example.json
+curl -X POST -F "file=@/path/to/file.pdf" http://localhost:3000/convert
 ```
 
-Convert a PDF file to text with a raw text response:
+Extract a password-protected PDF file's text content as JSON and save it to a file:
 ```bash
-curl -X POST -F "file=@/path/to/file.pdf" -F "raw=true" http://localhost:3000/convert -o example.txt
+curl -X POST -F "file=@/path/to/file.pdf" -F "password=XXX" -F "format=json" http://localhost:3000/convert -o example.json
 ```
 
+
+
 ## License
 
 This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.
diff --git a/main.mjs b/main.mjs
index 0615141..1269f74 100644
--- a/main.mjs
+++ b/main.mjs
@@ -49,8 +49,16 @@ app.post('/convert', upload.single('file'), async (req, res) => {
             }
         )();
 
-        // send the text
-        if (req.body.raw === "true") {
+        // send the content as raw text or JSON
+        if (String(req.body.format).toLowerCase() === "json") {
+            delete extractResult.filename;
+
+            extractResult.pages.forEach(page => {
+                page.content.forEach(content => delete content.fontName);
+            });
+
+            res.json(extractResult);
+        } else {
             res.send(
                 extractResult.pages.reduce(
                     (acc1, page) => acc1 + page.content.reduce(
@@ -60,14 +68,6 @@ app.post('/convert', upload.single('file'), async (req, res) => {
                     '',
                 ),
             );
-        } else {
-            delete extractResult.filename;
-
-            extractResult.pages.forEach(page => {
-                page.content.forEach(content => delete content.fontName);
-            });
-
-            res.json(extractResult);
         }
 
         // cleaning up