CSV DataStore better support

ahuarte47 · Oct 16, 2017 · 4208b8a · 4208b8a
1 parent 3c1b593
commit 4208b8a
Show file tree

Hide file tree

Showing 3 changed files with 61 additions and 0 deletions.
diff --git a/modules/library/main/src/main/java/org/geotools/util/CommonsConverterFactory.java b/modules/library/main/src/main/java/org/geotools/util/CommonsConverterFactory.java
@@ -237,6 +237,15 @@ public <T> T convert(Object source, Class<T> target) throws Exception {
             }
             catch( Exception ignore){
             }
+            for (String formatPattern : new String[]{ "yyyy-MM-dd HH:mm:ss", "yyyy-MM-ddTHH:mm:ss", "yyyy-MM-dd" }) {
+                try {
+                    SimpleDateFormat format3 = new SimpleDateFormat(formatPattern);
+                    Date parsed = format3.parse(string);
+                    return target.cast(parsed);
+                }
+                catch( Exception ignore){
+                }
+            }
             return null;
         }
     }

diff --git a/modules/unsupported/csv/src/main/java/org/geotools/data/csv/CSVDataStoreFactory.java b/modules/unsupported/csv/src/main/java/org/geotools/data/csv/CSVDataStoreFactory.java
@@ -157,6 +157,25 @@ private FileDataStore createDataStoreFromFile(File file, URI namespace,
         CSVFileState csvFileState = new CSVFileState(file, namespace);
         Object strategyParam = STRATEGYP.lookUp(params);
         CSVStrategy csvStrategy = null;
+
+        // Lookup geometry fields better.
+        if (strategyParam == null || params == null || params.size() == 0) {
+            org.geotools.feature.simple.SimpleFeatureTypeBuilder builder = CSVStrategy.createBuilder(csvFileState);
+
+            if (builder != null) {
+                org.opengis.feature.type.AttributeDescriptor lonDescr = builder.get("longitude");
+                org.opengis.feature.type.AttributeDescriptor latDescr = builder.get("latitude");
+                if (lonDescr == null || latDescr == null) {
+                    lonDescr = builder.get("lon");
+                    latDescr = builder.get("lat");
+                }
+                if (lonDescr != null && lonDescr.getType().getBinding().isAssignableFrom(Double.class) &&
+                    latDescr != null && latDescr.getType().getBinding().isAssignableFrom(Double.class)) {
+                    strategyParam = "guess";
+                }
+            }
+        }
+
         if (strategyParam != null) {
             String strategyString = strategyParam.toString();
             if (strategyString.equalsIgnoreCase("guess")) {

diff --git a/modules/unsupported/csv/src/main/java/org/geotools/data/csv/parse/CSVStrategy.java b/modules/unsupported/csv/src/main/java/org/geotools/data/csv/parse/CSVStrategy.java
@@ -108,6 +108,12 @@ public static SimpleFeatureTypeBuilder createBuilder(CSVFileState csvFileState,
     protected static Map<String, Class<?>> findMostSpecificTypesFromData(CsvReader csvReader,
             String[] headers) throws IOException {
         Map<String, Class<?>> result = new HashMap<String, Class<?>>();
+
+        List<java.time.format.DateTimeFormatter> formatterList = new java.util.ArrayList<java.time.format.DateTimeFormatter>();
+        formatterList.add(java.time.format.DateTimeFormatter.ISO_DATE_TIME);
+        formatterList.add(java.time.format.DateTimeFormatter.ISO_DATE);
+        int featureCount = 0;
+
         // start off assuming Integers for everything
         for (String header : headers) {
             result.put(header, Integer.class);
@@ -123,6 +129,29 @@ protected static Map<String, Class<?>> findMostSpecificTypesFromData(CsvReader c
             for (String value : values) {
                 String header = headers[i];
                 Class<?> type = result.get(header);
+
+                // Parse Boolean and Date/Time attribute types too.
+                if (value == null || value.length() == 0) {
+                    i++;
+                    continue;
+                }
+                if (type == Boolean.class && (!value.equalsIgnoreCase("true") && !value.equalsIgnoreCase("false"))) {
+                    type = String.class;
+                }
+                if (type == Integer.class) {
+                    if (value.equalsIgnoreCase("true") || value.equalsIgnoreCase("false")) type = Boolean.class;
+                }
+                if (type == Integer.class || type == java.util.Date.class) {
+                    for (int j = 0, jcount = formatterList.size(); j < jcount; j++) {
+                        try {
+                            formatterList.get(j).parse(value.replace(' ', 'T'));
+                            type = java.util.Date.class;
+                            break;
+                        } catch (Exception ex) {
+                        }
+                    }
+                }
+
                 // For each value in the row, ensure we can still parse it as the 
                 // defined type for this column; if not, make it more general
                 if (type == Integer.class) {
@@ -146,6 +175,10 @@ protected static Map<String, Class<?>> findMostSpecificTypesFromData(CsvReader c
                 result.put(header, type);
                 i++;
             }
+
+            // Skip to parse very-big CSV files.
+            if (featureCount>=100) break;
+            featureCount++;
         }
         return result;
     }