@xapp/arachne-web-service 1.36.1 → 1.37.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
|
@@ -4,7 +4,7 @@ import { ServicePageKind } from "./selectServicePages";
|
|
|
4
4
|
/**
|
|
5
5
|
* Bumped whenever the prompt, schema or normalizer changes, so stored extractions can be compared by version.
|
|
6
6
|
*/
|
|
7
|
-
export declare const EXTRACTOR_VERSION = "2026-10-
|
|
7
|
+
export declare const EXTRACTOR_VERSION = "2026-10-31.1";
|
|
8
8
|
/**
|
|
9
9
|
* Default cap on the page text sent to the LLM, in tokens (estimated at four characters a token).
|
|
10
10
|
*/
|
|
@@ -17,7 +17,7 @@ const ServiceTaxonomy_1 = require("../taxonomy/ServiceTaxonomy");
|
|
|
17
17
|
/**
|
|
18
18
|
* Bumped whenever the prompt, schema or normalizer changes, so stored extractions can be compared by version.
|
|
19
19
|
*/
|
|
20
|
-
exports.EXTRACTOR_VERSION = "2026-10-
|
|
20
|
+
exports.EXTRACTOR_VERSION = "2026-10-31.1";
|
|
21
21
|
/**
|
|
22
22
|
* Default cap on the page text sent to the LLM, in tokens (estimated at four characters a token).
|
|
23
23
|
*/
|
|
@@ -440,7 +440,7 @@ ${this.catalogText(catalog)}`;
|
|
|
440
440
|
|
|
441
441
|
Every value needs evidence: the URL of the page it is on and a snippet copied exactly, word for word, from that page (5 to 25 words). No evidence, no value. Do not infer, guess or generalise.
|
|
442
442
|
|
|
443
|
-
- services: catalog slugs for jobs the business says it performs.
|
|
443
|
+
- services: catalog slugs for jobs the business says it performs. List every specific job the pages name, one entry each: "drain cleaning", "sewer camera inspection" and "water heater replacement" are three jobs, each with its own slug, not one general plumbing slug. Read service lists, menus, headings, project descriptions and reviews for them. Use a slug marked (general) only for a service the site names in general terms, e.g. "AC repair", when no more specific job for it is named anywhere on the pages. Anything the business offers that has no catalog slug goes in unmapped_services, worded as the site words it.
|
|
444
444
|
- explicit_exclusions: anything the site says it does not do.
|
|
445
445
|
- flags: include a flag only when the pages state it. value is false only when the site says the opposite. Leave out flags the pages do not mention.
|
|
446
446
|
- attributes: certifications, equipment types, specialty materials, gear and sub-trade specialisms from the lists below that the pages name.
|
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@xapp/arachne-web-service",
|
|
3
|
-
"version": "1.
|
|
3
|
+
"version": "1.37.0",
|
|
4
4
|
"types": "lib/index",
|
|
5
5
|
"main": "lib/index",
|
|
6
6
|
"files": [
|
|
@@ -34,7 +34,7 @@
|
|
|
34
34
|
},
|
|
35
35
|
"dependencies": {
|
|
36
36
|
"@xapp/arachne": "1.18.0",
|
|
37
|
-
"@xapp/arachne-parsers": "1.27.
|
|
37
|
+
"@xapp/arachne-parsers": "1.27.1",
|
|
38
38
|
"google-auth-library": "11.3.2",
|
|
39
39
|
"gpt-3-encoder": "1.1.4",
|
|
40
40
|
"ico-to-png": "0.2.2",
|