Merge pull request #754 from OpenKnowledgeMaps/bugfixes-2024-02

Bugfixes 2024 02

Author: chreman

local_dev/config_local_headstart.ini

@@ -7,7 +7,7 @@ preprocessing_dir = "/var/www/html/dev/server/preprocessing/"
 # Full path to the images directory for the client. Needs to be in the public_html/www directory. Make sure that your webserver has write access to this directory.
 images_path = "/var/www/html/dev/server/paper_preview/"
 # Host of the client visualization
-host = "localhost/"
+host = "dev-searchflow-1/"
 # Relative path to the client visualization. Needs to be in the public_html/www directory.
 vis_path = "headstart"
 # Relative path to the client REST services. Needs to be in the public_html/www directory.
@@ -20,19 +20,21 @@ api_flavor = "dev"
 
 [snapshot]
 # Set to 1 to enable snapshot feature, 0 to disable
-snapshot_enabled = 0
+snapshot_enabled = 1
 # Absolute path to node binary
-node_path = "/path/to/node"
+node_path = "/usr/local/nvm/versions/node/v10.17.0/bin/node"
 # Absolute paths to node_modules folder, where puppeteer is installed
-nodemodules_path = "/path/to/node_modules"
+nodemodules_path = "/usr/local/nvm/versions/node/v10.17.0/lib/node_modules/"
 # Absolute path to getChartSVG.js
-getsvg_path = "/path/to/getChartSVG.js"
+getsvg_path = "/var/www/html/headstart/server/services/getChartSVG.js"
 # Absolute path to the directory, where the snapshots are stored. Webserver must have write access to this directory
-storage_path = "/path/to/storage/"
+storage_path = "/var/www/html/snapshots/"
 # PHP File responsible for rendering the bubble in a way to be snapshotted. Relative path to general host
-snapshot_php = "server/services/snapshot/headstart_snapshot.php"
+snapshot_php = "headstart/server/services/snapshot/headstart_snapshot.php"
 # snapshot_local_protocol fallback for non-server environments
 snapshot_local_protocol = "http://"
+# Size of thumbnail width
+snapshot_width = "1200px"
 
 [connection]
 # PostgreSQL database

local_dev/searchflow-container/Dockerfile

@@ -5,8 +5,19 @@ MAINTAINER Chris Kittel "christopher.kittel@openknowledgemaps.org"
 RUN a2enmod rewrite
 RUN apt-get update && apt-get install -y curl libsqlite3-dev php7.4-sqlite libonig-dev libxml2-dev
 RUN docker-php-ext-install pdo pdo_sqlite mbstring xml fileinfo
+# for puppeteer
+RUN apt-get install -y gconf-service libasound2 libatk1.0-0 libatk-bridge2.0-0 libc6 libcairo2 libcups2 libdbus-1-3 libexpat1 libfontconfig1 libgcc1 libgconf-2-4 libgdk-pixbuf2.0-0 libglib2.0-0 libgtk-3-0 libnspr4 libpango-1.0-0 libpangocairo-1.0-0 libstdc++6 libx11-6 libx11-xcb1 libxcb1 libxcomposite1 libxcursor1 libxdamage1 libxext6 libxfixes3 libxi6 libxrandr2 libxrender1 libxss1 libxtst6 ca-certificates fonts-liberation libappindicator1 libnss3 lsb-release xdg-utils wget
+ENV NVM_DIR /usr/local/nvm
+ENV NODE_VERSION 10.17.0
+RUN curl https://raw.githubusercontent.com/creationix/nvm/v0.24.1/install.sh | bash \
+    && . $NVM_DIR/nvm.sh \
+    && nvm install $NODE_VERSION \
+    && nvm alias default $NODE_VERSION \
+    && nvm use default
 
-RUN sed -i 's#AllowOverride [Nn]one#AllowOverride All#' /etc/apache2/apache2.conf
-RUN mkdir /var/www/localstorage
-COPY test.sqlite /var/www/localstorage/local.sqlite
-RUN chown -R www-data:www-data /var/www/localstorage
+ENV NODE_PATH $NVM_DIR/versions/node/v$NODE_VERSION/lib/node_modules
+ENV PATH      $NVM_DIR/versions/node/v$NODE_VERSION/bin:$PATH
+ENV NPM_PATH  $NVM_DIR/versions/node/v$NODE_VERSION/bin/npm
+RUN $NPM_PATH install -g puppeteer@^2.0.0 --unsafe-perm=true --allow-root
+
+RUN sed -i 's#AllowOverride [Nn]one#AllowOverride All#' /etc/apache2/apache2.conf

server/classes/headstart/preprocessing/Snapshot.php

@@ -10,22 +10,9 @@ public function __construct($ini_array, $query, $id, $service, $service_name, $v
             "query" => $query,
             "file" => $id,
             "service_name" => $service_name,
-            "service" => $service
+            "service" => $service,
+            "vis_type" => $vis_type
         );
-        if ($service_name == "LinkedCat") {
-          $post_data["vis_type"] = $vis_type;
-          $post_data["service_name"] = "linkedcat";
-          }
-        if ($service == "linkedcat_authorview") {
-          $post_data["vis_mode"] = "authors";
-          $post_data["service"] = "linkedcat";
-          $post_data["service_name"] = "linkedcat";
-        }
-        if ($service == "linkedcat_browseview") {
-          $post_data["vis_mode"] = "browse";
-          $post_data["service"] = "linkedcat";
-          $post_data["service_name"] = "linkedcat";
-        }
         $url_postfix = http_build_query($post_data);
 
         $node_path = $ini_array["snapshot"]["node_path"];

server/preprocessing/other-scripts/base.R

@@ -199,7 +199,7 @@ etl <- function(res, repo, non_public) {
   metadata$relation = check_metadata(res$dcrelation)
   metadata$identifier = check_metadata(res$dcidentifier)
   metadata$title = check_metadata(res$dctitle)
-  metadata$title = str_replace(metadata$title, " ...$", "")
+  metadata$title = gsub(" \\.\\.\\.$", "", metadata$title)
   metadata$paper_abstract = check_metadata(res$dcdescription)
   metadata$published_in = check_metadata(res$dcsource)
   metadata$year = check_metadata(res$dcdate)

server/preprocessing/other-scripts/openaire.R

@@ -144,6 +144,8 @@ preprocess_data <- function(all_artifacts){
                                                                              else {
                                                                                return (x)}
                                                                               })) # removes ;dk;atira;pure;researchoutput;pubmedpublicationtype;D013486
+   # remove greater / smaller than signs
+  all_artifacts$subject <- unlist(lapply(all_artifacts$subject, function(x) {gsub("&gt|&lt", "", x)}))
   all_artifacts$paper_abstract <- unlist(lapply(all_artifacts$paper_abstract, function(x){gsub("\\n", " ", x)}))
   all_artifacts$doi <- unlist(lapply(all_artifacts$doi, function(x) {str_replace_all(x, "[\r\n\t]" , "")}))
   return (all_artifacts)

server/preprocessing/other-scripts/test/params_base.json

@@ -7,5 +7,5 @@
   "min_descsize": 300,
   "limit": 120,
   "list_size": 100,
-  "custom_clustering": "mesh"
+  "repo": "ftunivbern"
 }

server/preprocessing/other-scripts/test/params_openaire.json

@@ -1,4 +1,4 @@
 {"funder":"EC",
- "project_id":"267439",
+ "project_id":"643410",
  "list_size": -1,
  "vis_id": "TEST_ID"}

server/preprocessing/other-scripts/test/test_base.R

@@ -7,7 +7,7 @@ options(warn=1)
 wd <- dirname(dirname(rstudioapi::getActiveDocumentContext()$path))
 setwd(wd) #Don't forget to set your working directory
 
-query <- "species" #args[2]
+query <- "parzival" #args[2]
 service <- "base"
 params <- NULL
 params_file <- "test/params_base.json"

server/preprocessing/other-scripts/test/test_openaire.R

@@ -9,7 +9,7 @@ setwd(wd) #Don't forget to set your working directory
 
 # query: project acryonm
 # params: grant_id and funding_stream
-query <- "GRINDOOR" #args[2]
+query <- "openaire2020" #args[2]
 service <- "openaire"
 params <- NULL
 params_file <- "test/params_openaire.json"
@@ -71,4 +71,5 @@ if (!exists('output_json')) {
   output_json <- enrich_output_json(output_json)
 }
 
-print(output_json)
+#print(output_json)
+output <- fromJSON(output_json)

server/services/getChartSVG.js

@@ -10,7 +10,7 @@ function timeout(ms) {
     const page = await browser.newPage();
     await page.setViewport({width: 1920, height: 1080})
     await page.goto(process.argv[2], {waitUntil: 'networkidle2'});
-    await timeout(2000)
+    await timeout(3000)
     await page.screenshot({path: process.argv[3], clip: { x: 0, y: 0, width: 1150, height: 1080 }});
     browser.close();