+ )
+
+}
+
+export default NERContext;
diff --git a/src/main/js/frontend/src/SideBarData.js b/src/main/js/frontend/src/SideBarData.js
index 141058a..a4704b0 100644
--- a/src/main/js/frontend/src/SideBarData.js
+++ b/src/main/js/frontend/src/SideBarData.js
@@ -7,6 +7,10 @@ export const SideBarData = [
icon: ,
key: "/"
},
+ {
+ label: "NER",
+ key: "/ner"
+ },
{
label: "Loading",
icon: ,
diff --git a/src/main/xar-resources/modules/rest-api.xqm b/src/main/xar-resources/modules/rest-api.xqm
index a252cfe..fc04fe5 100644
--- a/src/main/xar-resources/modules/rest-api.xqm
+++ b/src/main/xar-resources/modules/rest-api.xqm
@@ -1,6 +1,7 @@
xquery version "3.1";
module namespace api = "http://exist-db.org/xquery/stanford-nlp/api";
+import module namespace ner = "http://exist-db.org/xquery/stanford-nlp/ner";
import module namespace scheduler = "http://exist-db.org/xquery/scheduler";
import module namespace map = "http://www.w3.org/2005/xpath-functions/map";
@@ -131,3 +132,37 @@ function api:logs($timestamp as xs:string*) as map(*)
}
}
};
+
+(:~
+ : This method runs the ner:clasify($text, $properties) on the text passed in for the language specified.
+ : @param $content the properties of the source graph in a JSON object
+ : @see ner:properties-from-language()
+ : @return A map
+ : @custom:openapi-tag Natural Language Processing
+ :)
+declare
+ %rest:POST("{$content}")
+ %rest:path("/Stanford/ner")
+ %rest:consumes("application/json")
+ %rest:produces("application/json")
+ %output:media-type("application/json")
+ %output:method("json")
+function api:query-text-as-json($content as xs:string) as map(*) {
+ let $postBody := fn:parse-json(util:base64-decode($content))
+ let $properties := ner:properties-from-language($postBody?language)
+ let $classified := ner:classify($postBody?text, $properties)
+ return
+ try {
+ map {
+ 'text' : ner:stringify($classified)
+ }
+ } catch * {
+ map {
+ 'code': $err:code,
+ 'description': $err:description,
+ 'value': $err:value,
+ 'properties': $properties
+ }
+ }
+};
+
diff --git a/src/main/xquery/ner-module.xqm b/src/main/xquery/ner-module.xqm
index 1f0b4b1..75a69cc 100644
--- a/src/main/xquery/ner-module.xqm
+++ b/src/main/xquery/ner-module.xqm
@@ -11,22 +11,26 @@ import module namespace functx = "http://www.functx.com";
import module namespace rest = "http://exquery.org/ns/restxq";
+declare namespace output = "http://www.w3.org/2010/xslt-xquery-serialization";
+declare namespace array = "http://www.w3.org/2005/xpath-functions/array";
+declare namespace map = "http://www.w3.org/2005/xpath-functions/map";
+
(:~
: Generates the snippet match string to show the highlighted text for the client.
:
: @param $match The match text for a snippet that contains highlighted text
: @return A string with highlight spans encoded within the string
:)
-declare function ner:stringify($match as node()) as xs:string {
+declare function ner:stringify($match) as xs:string {
fn:string-join(
- for $text-or-highlight in $match/node()
+ for $text-or-highlight in $match
return
- if ($text-or-highlight instance of element())
+ if ($text-or-highlight instance of element())
then
fn:concat('',
- $text-or-highlight/text(),
+ '">',
+ $text-or-highlight/text(),
'')
else
$text-or-highlight
@@ -35,7 +39,7 @@ declare function ner:stringify($match as node()) as xs:string {
(:~
- : This method runs the ner:clasify($text, $properties) on the leaf nodes of the
+ : This method runs the ner:clasify($text, $properties) on the leaf nodes of the
: document for the defaults for the English language. The english defaults are
: in the file /db/apps/stanford-nlp/data/StanfordCoreNLP-english.json
: @param $request-body The document to process.
@@ -46,7 +50,7 @@ function ner:classify-document($request-body as document-node(element())) as nod
};
(:~
- : This method runs the ner:clasify($text, $properties) on the leaf nodes of the
+ : This method runs the ner:clasify($text, $properties) on the leaf nodes of the
: document for the language specified.
: @param $request-body The document to process.
: @param $language The two character string to identify the language to process.
@@ -58,7 +62,7 @@ function ner:classify-document($request-body as document-node(element()), $langu
};
(:~
- : This method runs the ner:clasify($text, $properties) on the leaf nodes of the
+ : This method runs the ner:clasify($text, $properties) on the leaf nodes of the
: XML identitied by $node for the defaults for the English language. The english defaults are
: in the file /db/apps/stanford-nlp/data/StanfordCoreNLP-english.json
: @param $node The node to process.
@@ -69,7 +73,7 @@ function ner:classify-node($node as node()) as node() {
};
(:~
- : This method runs the ner:clasify($text, $properties) on the leaf nodes of the
+ : This method runs the ner:clasify($text, $properties) on the leaf nodes of the
: node for the language specified.
: @param $node The node to process.
: @param $language The two character string to identify the language to process.
@@ -91,20 +95,16 @@ function ner:classify-node($node as node(), $language as xs:string) as node() {
: Any other value loads the english defaults.
: @param $language The two character string to identify the language for the default JSON document.
:)
-declare
+declare
function ner:properties-from-language($language as xs:string) as map(*) {
- try {
- switch ($language)
- case "en" return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-english.json")
- case "ar" return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-arabic.json")
- case "es" return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-spanish.json")
- case "fr" return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-french.json")
- case "zh" return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-chinese.json")
- case "de" return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-german.json")
- default return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-english.json")
- } catch * {
- fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-english.json")
- }
+ switch ($language)
+ case "en" return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-english.json")
+ case "ar" return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-arabic.json")
+ case "es" return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-spanish.json")
+ case "fr" return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-french.json")
+ case "zh" return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-chinese.json")
+ case "de" return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-german.json")
+ default return fn:json-doc("/db/apps/stanford-nlp/data/StanfordCoreNLP-english.json")
};
(:~
@@ -113,6 +113,7 @@ function ner:properties-from-language($language as xs:string) as map(*) {
: @param $language The two character string to identify the language to process.
: @see ner:properties-from-language()
: @return An XML node
+ : @custom:openapi-tag Natural Language Processing
:)
declare
%rest:GET
@@ -121,36 +122,11 @@ declare
%rest:query-param("lang", "{$language}", "en")
%rest:produces("application/xml")
function ner:query-text-as-xml($text as xs:string*, $language as xs:string*) as node() {
- element { 'ner' } {
+ element { 'ner' } {
ner:classify(
- util:unescape-uri($text[1], "UTF-8"),
+ util:unescape-uri($text[1], "UTF-8"),
ner:properties-from-language($language[1])
- )
- }
-};
-
-(:~
- : This method runs the ner:clasify($text, $properties) on the text passed in for the language specified.
- : @param $text The text to process.
- : @param $language The two character string to identify the language to process.
- : @see ner:properties-from-language()
- : @return A map
- :)
-declare
- %rest:GET
- %rest:path("/StanfordNLP/NER")
- %rest:query-param("text", "{$text}")
- %rest:query-param("lang", "{$language}", "en")
- %rest:produces("application/json")
-function ner:query-text-as-json($text as xs:string*, $language as xs:string*) as map(*) {
- map {
- 'text' :
- ner:stringify(
- ner:classify(
- util:unescape-uri($text[1], "UTF-8"),
- ner:properties-from-language($language[1])
- )
- )
+ )
}
};
@@ -189,14 +165,14 @@ declare function ner:enrich($text as xs:string, $tokens as node()*) {
let $ner-text := fn:substring($text, $start, $length)
let $next := fn:subsequence($tokens, fn:index-of($tokens, $sibling-token) + 1)
return (
- ner:enrich($before, $next),
- element { $last-token/NER/text() } { $ner-text },
+ ner:enrich($before, $next),
+ element { $last-token/NER/text() } { $ner-text },
if (fn:string-length($after) gt 0) then $after else ())
-
+
};
(:~
- : This method runs the ner:clasify($text, $properties) on the leaf nodes of the
+ : This method runs the ner:clasify($text, $properties) on the leaf nodes of the
: node for the properties of the language.
: @param $node The node to process.
: @param $properties The map of the values for proceessing the pipeline.
@@ -215,9 +191,9 @@ declare function ner:dispatch($node as node()?, $properties as map(*)) {
:)
declare function ner:pass-through($node as node()?, $properties as map(*)) {
if ($node)
- then element { $node/name() } {
- $node/@*,
- for $cnode in $node/*
+ then element { $node/name() } {
+ $node/@*,
+ for $cnode in $node/*
return ner:dispatch($cnode, $properties)
}
else ()
@@ -230,5 +206,5 @@ let $tokens := for $token in nlp:parse($text, $properties)//token[fn:not(NER = "
let $token-start := $token/CharacterOffsetBegin/number()
order by $token-start descending
return $token
-return ner:enrich($text, $tokens)
+return ner:enrich($text, $tokens)
};
diff --git a/xar-assembly.xml b/xar-assembly.xml
index f1128b6..05a983f 100644
--- a/xar-assembly.xml
+++ b/xar-assembly.xml
@@ -24,7 +24,7 @@
pre-install.xqpost-install.xq
-
+