New Components - docparser

components/docparser/actions/fetch-document-url/fetch-document-url.mjs

@@ -0,0 +1,38 @@
+import FormData from "form-data";
+import docparser from "../../docparser.app.mjs";
+
+export default {
+  key: "docparser-fetch-document-url",
+  name: "Fetch Document by URL",
+  description: "Fetches a document from a provided URL and imports it to Docparser for parsing. [See the documentation](https://docparser.com/api/)",
+  version: "0.0.1",
+  type: "action",
+  props: {
+    docparser,
+    parserId: {
+      propDefinition: [
+        docparser,
+        "parserId",
+      ],
+    },
+    url: {
+      type: "string",
+      label: "Document URL",
+      description: "The URL of the document to be fetched and imported into Docparser.",
+    },
+  },
+  async run({ $ }) {
+    const data = new FormData();
+    data.append("url", this.url);
+
+    const response = await this.docparser.fetchDocumentFromURL({
+      $,
+      parserId: this.parserId,
+      data,
+      headers: data.getHeaders(),
+    });
+
+    $.export("$summary", `Document is scheduled to be fetched and processed. Document ID: ${response.document_id}`);
+    return response;
+  },
+};

components/docparser/actions/upload-document/upload-document.mjs

@@ -0,0 +1,40 @@
+import FormData from "form-data";
+import fs from "fs";
+import { checkTmp } from "../../common/utils.mjs";
+import docparser from "../../docparser.app.mjs";
+
+export default {
+  key: "docparser-upload-document",
+  name: "Upload Document",
+  description: "Uploads a document to docparser that initiates parsing immediately after reception. [See the documentation](https://docparser.com/api/#import-documents)",
+  version: "0.0.1",
+  type: "action",
+  props: {
+    docparser,
+    parserId: {
+      propDefinition: [
+        docparser,
+        "parserId",
+      ],
+    },
+    file: {
+      type: "string",
+      label: "File",
+      description: "The path to a file in the `/tmp` directory. [See the documentation on working with files](https://pipedream.com/docs/code/nodejs/working-with-files/#writing-a-file-to-tmp)",
+    },
+  },
+  async run({ $ }) {
+    const data = new FormData();
+    data.append("file", fs.createReadStream(checkTmp(this.file)));
+
+    const response = await this.docparser.uploadDocument({
+      $,
+      parserId: this.parserId,
+      data,
+      headers: data.getHeaders(),
+    });
+
+    $.export("$summary", `Successfully uploaded document. Document ID: ${response.id}`);
+    return response;
+  },
+};

components/docparser/common/utils.mjs

@@ -0,0 +1,6 @@
+export const checkTmp = (filename) => {
+  if (!filename.startsWith("/tmp")) {
+    return `/tmp/${filename}`;
+  }
+  return filename;
+};

components/docparser/docparser.app.mjs

@@ -1,11 +1,73 @@
+import { axios } from "@pipedream/platform";
+
 export default {
   type: "app",
   app: "docparser",
-  propDefinitions: {},
+  propDefinitions: {
+    parserId: {
+      type: "string",
+      label: "Parser ID",
+      description: "The ID of the parser to be used.",
+      async options() {
+        const parsers = await this.listParsers();
+        return parsers.map(({
+          id: value, label,
+        }) => ({
+          label,
+          value,
+        }));
+      },
+    },
+  },
   methods: {
-    // this.$auth contains connected account data
-    authKeys() {
-      console.log(Object.keys(this.$auth));
+    _baseUrl() {
+      return "https://api.docparser.com";
+    },
+    _auth() {
+      return {
+        username: `${this.$auth.api_key}`,
+        password: "",
+      };
+    },
+    _makeRequest({
+      $ = this, path, ...opts
+    }) {
+      return axios($, {
+        url: this._baseUrl() + path,
+        auth: this._auth(),
+        ...opts,
+      });
+    },
+    listData({
+      parserId, ...opts
+    }) {
+      return this._makeRequest({
+        path: `/v1/results/${parserId}`,
+        ...opts,
+      });
+    },
+    listParsers() {
+      return this._makeRequest({
+        path: "/v1/parsers",
+      });
+    },
+    fetchDocumentFromURL({
+      parserId, ...opts
+    }) {
+      return this._makeRequest({
+        method: "POST",
+        path: `/v2/document/fetch/${parserId}`,
+        ...opts,
+      });
+    },
+    uploadDocument({
+      parserId, ...opts
+    }) {
+      return this._makeRequest({
+        method: "POST",
+        path: `/v1/document/upload/${parserId}`,
+        ...opts,
+      });
     },
   },
 };

components/docparser/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@pipedream/docparser",
-  "version": "0.6.0",
+  "version": "0.1.0",
   "description": "Pipedream docparser Components",
   "main": "docparser.app.mjs",
   "keywords": [
@@ -13,6 +13,6 @@
     "access": "public"
   },
   "dependencies": {
-    "@pipedream/platform": "^3.0.0"
+    "@pipedream/platform": "^3.0.3"
   }
 }

components/docparser/sources/common/base.mjs

@@ -0,0 +1,69 @@
+import { DEFAULT_POLLING_SOURCE_TIMER_INTERVAL } from "@pipedream/platform";
+import app from "../../docparser.app.mjs";
+
+export default {
+  props: {
+    app,
+    db: "$.service.db",
+    timer: {
+      type: "$.interface.timer",
+      default: {
+        intervalSeconds: DEFAULT_POLLING_SOURCE_TIMER_INTERVAL,
+      },
+    },
+    parserId: {
+      propDefinition: [
+        app,
+        "parserId",
+      ],
+    },
+  },
+  methods: {
+    _getLastDate() {
+      return this.db.get("lastDate") || "1970-01-01T00:00:00";
+    },
+    _setLastDate(lastDate) {
+      this.db.set("lastDate", lastDate);
+    },
+    async emitEvent(maxResults = false) {
+      const lastDate = this._getLastDate();
+      const fn = this.getFunction();
+      const params = {
+        sort_by: "parsed_at",
+        sort_order: "DESC",
+        list: "processed_after",
+        date: lastDate,
+      };
+
+      if (maxResults) {
+        params.limit = maxResults;
+      }
+
+      const response = await fn({
+        parserId: this.parserId,
+        params,
+      });
+
+      if (response.length) {
+        const dateTime = response[0].processed_at_utc;
+        this._setLastDate(dateTime.substring(0, dateTime.length - 6));
+      }
+
+      for (const item of response.reverse()) {
+        this.$emit(item, {
+          id: item.id,
+          summary: this.getSummary(item),
+          ts: Date.parse(item.created || new Date()),
+        });
+      }
+    },
+  },
+  hooks: {
+    async deploy() {
+      await this.emitEvent(25);
+    },
+  },
+  async run() {
+    await this.emitEvent();
+  },
+};

components/docparser/sources/new-document-data-available/new-document-data-available.mjs

@@ -0,0 +1,22 @@
+import common from "../common/base.mjs";
+import sampleEmit from "./test-event.mjs";
+
+export default {
+  ...common,
+  key: "docparser-new-document-data-available",
+  name: "New Document Data Available",
+  description: "Emit new event every time a document is processed and parsed data is available. [See the documentation](https://docparser.com/api/)",
+  version: "0.0.1",
+  type: "source",
+  dedupe: "unique",
+  methods: {
+    ...common.methods,
+    getFunction() {
+      return this.app.listData;
+    },
+    getSummary(item) {
+      return `New Document Parsed: ${item.file_name}`;
+    },
+  },
+  sampleEmit,
+};

components/docparser/sources/new-document-data-available/test-event.mjs

@@ -0,0 +1,16 @@
+export default {
+  "id": "2b11659f161dcd26694e9403fc430cfd",
+  "document_id": "2b11659f161dcd26694e9403fc430cfd",
+  "remote_id": "",
+  "file_name": "file.pdf",
+  "media_link": "https://api.docparser.com/v1/document/media/2b11659f161dcd26694e9403fc430cfd-2b11659f161dcd26694e9403fc430cfd",
+  "media_link_original": "https://api.docparser.com/v1/document/media/2b11659f161dcd26694e9403fc430cfd-2b11659f161dcd26694e9403fc430cfd/original",
+  "media_link_data": "https://api.docparser.com/v1/document/media/2b11659f161dcd26694e9403fc430cfd-2b11659f161dcd26694e9403fc430cfd/data",
+  "page_count": 5,
+  "uploaded_at": "2025-04-08T13:32:02+00:00",
+  "processed_at": "2025-04-08T13:32:02+00:00",
+  "uploaded_at_utc": "2025-04-08T13:32:02+00:00",
+  "uploaded_at_user": "2025-04-08T06:32:02+00:00",
+  "processed_at_utc": "2025-04-08T13:32:02+00:00",
+  "processed_at_user": "2025-04-08T06:32:02+00:00"
+}