embulk-parser-poi_excel 0.1.4 → 0.1.11

data/src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_cellType.java

@@ -0,0 +1,79 @@
+package org.embulk.parser.poi_excel;
+
+import static org.hamcrest.CoreMatchers.is;
+import static org.hamcrest.MatcherAssert.assertThat;
+
+import java.net.URL;
+import java.text.ParseException;
+import java.util.Arrays;
+import java.util.List;
+
+import org.apache.poi.ss.usermodel.CellType;
+import org.embulk.parser.EmbulkPluginTester;
+import org.embulk.parser.EmbulkTestOutputPlugin.OutputRecord;
+import org.embulk.parser.EmbulkTestParserConfig;
+import org.junit.experimental.theories.DataPoints;
+import org.junit.experimental.theories.Theories;
+import org.junit.experimental.theories.Theory;
+import org.junit.runner.RunWith;
+
+@RunWith(Theories.class)
+public class TestPoiExcelParserPlugin_cellType {
+
+	@DataPoints
+	public static String[] FILES = { "test1.xls", "test2.xlsx" };
+
+	@Theory
+	public void testCellType(String excelFile) throws ParseException {
+		try (EmbulkPluginTester tester = new EmbulkPluginTester()) {
+			tester.addParserPlugin(PoiExcelParserPlugin.TYPE, PoiExcelParserPlugin.class);
+
+			EmbulkTestParserConfig parser = tester.newParserConfig(PoiExcelParserPlugin.TYPE);
+			parser.set("sheets", Arrays.asList("test1"));
+			parser.set("skip_header_lines", 3);
+			parser.addColumn("long", "long").set("column_number", "A").set("value", "cell_type");
+			parser.addColumn("string", "string").set("column_number", "A").set("value", "cell_type");
+
+			URL inFile = getClass().getResource(excelFile);
+			List<OutputRecord> result = tester.runParser(inFile, parser);
+
+			assertThat(result.size(), is(5));
+			check1(result, 0, CellType.NUMERIC, "NUMERIC");
+			check1(result, 1, CellType.STRING, "STRING");
+			check1(result, 2, CellType.FORMULA, "FORMULA");
+			check1(result, 3, CellType.BOOLEAN, "BOOLEAN");
+			check1(result, 4, CellType.FORMULA, "FORMULA");
+		}
+	}
+
+	@Theory
+	public void testCellCachedType(String excelFile) throws ParseException {
+		try (EmbulkPluginTester tester = new EmbulkPluginTester()) {
+			tester.addParserPlugin(PoiExcelParserPlugin.TYPE, PoiExcelParserPlugin.class);
+
+			EmbulkTestParserConfig parser = tester.newParserConfig(PoiExcelParserPlugin.TYPE);
+			parser.set("sheets", Arrays.asList("test1"));
+			parser.set("skip_header_lines", 3);
+			parser.addColumn("long", "long").set("column_number", "A").set("value", "cell_cached_type");
+			parser.addColumn("string", "string").set("column_number", "A").set("value", "cell_cached_type");
+
+			URL inFile = getClass().getResource(excelFile);
+			List<OutputRecord> result = tester.runParser(inFile, parser);
+
+			assertThat(result.size(), is(5));
+			check1(result, 0, CellType.NUMERIC, "NUMERIC");
+			check1(result, 1, CellType.STRING, "STRING");
+			check1(result, 2, CellType.BOOLEAN, "BOOLEAN");
+			check1(result, 3, CellType.BOOLEAN, "BOOLEAN");
+			check1(result, 4, CellType.ERROR, "ERROR");
+		}
+	}
+
+	@SuppressWarnings("deprecation")
+	private void check1(List<OutputRecord> result, int index, CellType cellType, String s) throws ParseException {
+		OutputRecord r = result.get(index);
+		// System.out.println(r);
+		assertThat(r.getAsLong("long"), is((long) cellType.getCode()));
+		assertThat(r.getAsString("string"), is(s));
+	}
+}

data/src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_columnNumber.java

@@ -2,7 +2,7 @@ package org.embulk.parser.poi_excel;
 
 import static org.hamcrest.CoreMatchers.is;
 import static org.hamcrest.CoreMatchers.nullValue;
-import static org.junit.Assert.assertThat;
+import static org.hamcrest.MatcherAssert.assertThat;
 
 import java.net.URL;
 import java.text.ParseException;

data/src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_constant.java

@@ -2,7 +2,7 @@ package org.embulk.parser.poi_excel;
 
 import static org.hamcrest.CoreMatchers.is;
 import static org.hamcrest.CoreMatchers.nullValue;
-import static org.junit.Assert.assertThat;
+import static org.hamcrest.MatcherAssert.assertThat;
 
 import java.net.URL;
 import java.text.ParseException;

data/src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_convertError.java

@@ -2,7 +2,7 @@ package org.embulk.parser.poi_excel;
 
 import static org.hamcrest.CoreMatchers.is;
 import static org.hamcrest.CoreMatchers.nullValue;
-import static org.junit.Assert.assertThat;
+import static org.hamcrest.MatcherAssert.assertThat;
 import static org.junit.Assert.fail;
 
 import java.net.URL;

data/src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_formula.java

@@ -0,0 +1,90 @@
+package org.embulk.parser.poi_excel;
+
+import static org.hamcrest.CoreMatchers.is;
+import static org.hamcrest.MatcherAssert.assertThat;
+
+import java.net.URL;
+import java.text.ParseException;
+import java.util.Arrays;
+import java.util.List;
+
+import org.embulk.config.ConfigSource;
+import org.embulk.parser.EmbulkPluginTester;
+import org.embulk.parser.EmbulkTestOutputPlugin.OutputRecord;
+import org.embulk.parser.EmbulkTestParserConfig;
+import org.junit.experimental.theories.DataPoints;
+import org.junit.experimental.theories.Theories;
+import org.junit.experimental.theories.Theory;
+import org.junit.runner.RunWith;
+
+@RunWith(Theories.class)
+public class TestPoiExcelParserPlugin_formula {
+
+	@DataPoints
+	public static String[] FILES = { "test1.xls", "test2.xlsx" };
+
+	@Theory
+	public void testForumlaHandlingCashedValue(String excelFile) throws ParseException {
+		try (EmbulkPluginTester tester = new EmbulkPluginTester()) {
+			tester.addParserPlugin(PoiExcelParserPlugin.TYPE, PoiExcelParserPlugin.class);
+
+			EmbulkTestParserConfig parser = tester.newParserConfig(PoiExcelParserPlugin.TYPE);
+			parser.set("sheet", "formula_replace");
+
+			parser.addColumn("text", "string").set("formula_handling", "cashed_value");
+
+			URL inFile = getClass().getResource(excelFile);
+			List<OutputRecord> result = tester.runParser(inFile, parser);
+
+			assertThat(result.size(), is(2));
+			assertThat(result.get(0).getAsString("text"), is("boolean"));
+			assertThat(result.get(1).getAsString("text"), is("test2-b1"));
+		}
+	}
+
+	@Theory
+	public void testForumlaHandlingEvaluate(String excelFile) throws ParseException {
+		try (EmbulkPluginTester tester = new EmbulkPluginTester()) {
+			tester.addParserPlugin(PoiExcelParserPlugin.TYPE, PoiExcelParserPlugin.class);
+
+			EmbulkTestParserConfig parser = tester.newParserConfig(PoiExcelParserPlugin.TYPE);
+			parser.set("sheet", "formula_replace");
+
+			parser.addColumn("text", "string").set("formula_handling", "evaluate");
+
+			URL inFile = getClass().getResource(excelFile);
+			List<OutputRecord> result = tester.runParser(inFile, parser);
+
+			assertThat(result.size(), is(2));
+			assertThat(result.get(0).getAsString("text"), is("boolean"));
+			assertThat(result.get(1).getAsString("text"), is("test2-b1"));
+		}
+	}
+
+	@Theory
+	public void testForumlaReplace(String excelFile) throws ParseException {
+		try (EmbulkPluginTester tester = new EmbulkPluginTester()) {
+			tester.addParserPlugin(PoiExcelParserPlugin.TYPE, PoiExcelParserPlugin.class);
+
+			EmbulkTestParserConfig parser = tester.newParserConfig(PoiExcelParserPlugin.TYPE);
+			parser.set("sheet", "formula_replace");
+
+			ConfigSource replace0 = tester.newConfigSource();
+			replace0.set("regex", "test1");
+			replace0.set("to", "merged_cell");
+			ConfigSource replace1 = tester.newConfigSource();
+			replace1.set("regex", "B1");
+			replace1.set("to", "B${row}");
+			parser.set("formula_replace", Arrays.asList(replace0, replace1));
+
+			parser.addColumn("text", "string");
+
+			URL inFile = getClass().getResource(excelFile);
+			List<OutputRecord> result = tester.runParser(inFile, parser);
+
+			assertThat(result.size(), is(2));
+			assertThat(result.get(0).getAsString("text"), is("test3-a1"));
+			assertThat(result.get(1).getAsString("text"), is("test2-b2"));
+		}
+	}
+}

data/src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_mergedCell.java

@@ -0,0 +1,94 @@
+package org.embulk.parser.poi_excel;
+
+import static org.hamcrest.CoreMatchers.is;
+import static org.hamcrest.MatcherAssert.assertThat;
+
+import java.net.URL;
+import java.text.ParseException;
+import java.util.List;
+
+import org.embulk.parser.EmbulkPluginTester;
+import org.embulk.parser.EmbulkTestOutputPlugin.OutputRecord;
+import org.embulk.parser.EmbulkTestParserConfig;
+import org.junit.experimental.theories.DataPoints;
+import org.junit.experimental.theories.Theories;
+import org.junit.experimental.theories.Theory;
+import org.junit.runner.RunWith;
+
+@RunWith(Theories.class)
+public class TestPoiExcelParserPlugin_mergedCell {
+
+	@DataPoints
+	public static String[] FILES = { "test1.xls", "test2.xlsx" };
+
+	@Theory
+	public void testSearchMergedCell_default(String excelFile) throws ParseException {
+		test(excelFile, null, true);
+	}
+
+	@Theory
+	public void testSearchMergedCell_true(String excelFile) throws ParseException {
+		// compatibility ver 0.1.7
+		test(excelFile, true, true);
+	}
+
+	@Theory
+	public void testSearchMergedCell_false(String excelFile) throws ParseException {
+		// compatibility ver 0.1.7
+		test(excelFile, false, false);
+	}
+
+	@Theory
+	public void testSearchMergedCell_none(String excelFile) throws ParseException {
+		test(excelFile, "none", false);
+	}
+
+	@Theory
+	public void testSearchMergedCell_linear(String excelFile) throws ParseException {
+		test(excelFile, "linear_search", true);
+	}
+
+	@Theory
+	public void testSearchMergedCell_tree(String excelFile) throws ParseException {
+		test(excelFile, "tree_search", true);
+	}
+
+	@Theory
+	public void testSearchMergedCell_hash(String excelFile) throws ParseException {
+		test(excelFile, "hash_search", true);
+	}
+
+	private void test(String excelFile, Object arg, boolean search) {
+		try (EmbulkPluginTester tester = new EmbulkPluginTester()) {
+			tester.addParserPlugin(PoiExcelParserPlugin.TYPE, PoiExcelParserPlugin.class);
+
+			EmbulkTestParserConfig parser = tester.newParserConfig(PoiExcelParserPlugin.TYPE);
+			parser.set("sheet", "merged_cell");
+			if (arg != null) {
+				parser.set("search_merged_cell", arg);
+			}
+			parser.addColumn("a", "string");
+			parser.addColumn("b", "string");
+
+			URL inFile = getClass().getResource(excelFile);
+			List<OutputRecord> result = tester.runParser(inFile, parser);
+
+			assertThat(result.size(), is(4));
+			if (search) {
+				check6(result, 0, "test3-a1", "test3-a1");
+			} else {
+				check6(result, 0, "test3-a1", null);
+			}
+			check6(result, 1, "data", "0");
+			check6(result, 2, null, null);
+			check6(result, 3, null, null);
+		}
+	}
+
+	private void check6(List<OutputRecord> result, int index, String a, String b) {
+		OutputRecord r = result.get(index);
+		// System.out.println(r);
+		assertThat(r.getAsString("a"), is(a));
+		assertThat(r.getAsString("b"), is(b));
+	}
+}

data/src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_sheets.java

@@ -1,7 +1,7 @@
 package org.embulk.parser.poi_excel;
 
 import static org.hamcrest.CoreMatchers.is;
-import static org.junit.Assert.assertThat;
+import static org.hamcrest.MatcherAssert.assertThat;
 
 import java.net.URL;
 import java.text.ParseException;
@@ -89,4 +89,38 @@ public class TestPoiExcelParserPlugin_sheets {
 		assertThat(record.getAsString("text"), is(text));
 		assertThat(record.getAsLong("number"), is(number));
 	}
+
+	@Theory
+	public void testResolveSheetName1(String excelFile) throws ParseException {
+		try (EmbulkPluginTester tester = new EmbulkPluginTester()) {
+			tester.addParserPlugin(PoiExcelParserPlugin.TYPE, PoiExcelParserPlugin.class);
+
+			EmbulkTestParserConfig parser = tester.newParserConfig(PoiExcelParserPlugin.TYPE);
+			parser.set("sheets", Arrays.asList("*es*"));
+			parser.addColumn("name", "string").set("value", "sheet_name");
+
+			URL inFile = getClass().getResource(excelFile);
+			List<OutputRecord> result = tester.runParser(inFile, parser);
+
+			OutputRecord record = result.get(0);
+			assertThat(record.getAsString("name"), is("test1"));
+		}
+	}
+
+	@Theory
+	public void testResolveSheetName2(String excelFile) throws ParseException {
+		try (EmbulkPluginTester tester = new EmbulkPluginTester()) {
+			tester.addParserPlugin(PoiExcelParserPlugin.TYPE, PoiExcelParserPlugin.class);
+
+			EmbulkTestParserConfig parser = tester.newParserConfig(PoiExcelParserPlugin.TYPE);
+			parser.set("sheets", Arrays.asList("test?"));
+			parser.addColumn("name", "string").set("value", "sheet_name");
+
+			URL inFile = getClass().getResource(excelFile);
+			List<OutputRecord> result = tester.runParser(inFile, parser);
+
+			OutputRecord record = result.get(0);
+			assertThat(record.getAsString("name"), is("test1"));
+		}
+	}
 }

metadata

@@ -1,19 +1,19 @@
 --- !ruby/object:Gem::Specification
 name: embulk-parser-poi_excel
 version: !ruby/object:Gem::Version
-  version: 0.1.4
+  version: 0.1.11
 platform: ruby
 authors:
 - hishidama
 autorequire:
 bindir: bin
 cert_chain: []
-date: 2015-11-10 00:00:00.000000000 Z
+date: 2020-09-13 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.0'
   name: bundler
@@ -21,13 +21,13 @@ dependencies:
   type: :development
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ~>
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '1.0'
 - !ruby/object:Gem::Dependency
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '10.0'
   name: rake
@@ -35,7 +35,7 @@ dependencies:
   type: :development
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - '>='
+    - - ">="
       - !ruby/object:Gem::Version
         version: '10.0'
 description: Parses Excel files(xls, xlsx) read by other file input plugins.
@@ -45,10 +45,19 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
-- .gitignore
+- ".gitignore"
 - LICENSE.txt
 - README.md
 - build.gradle
+- classpath/commons-codec-1.10.jar
+- classpath/commons-collections4-4.1.jar
+- classpath/curvesapi-1.04.jar
+- classpath/embulk-parser-poi_excel-0.1.11.jar
+- classpath/embulk-standards-0.7.5.jar
+- classpath/poi-3.17.jar
+- classpath/poi-ooxml-3.17.jar
+- classpath/poi-ooxml-schemas-3.17.jar
+- classpath/xmlbeans-2.6.0.jar
 - gradle/wrapper/gradle-wrapper.jar
 - gradle/wrapper/gradle-wrapper.properties
 - gradlew
@@ -60,10 +69,13 @@ files:
 - src/main/java/org/embulk/parser/poi_excel/bean/PoiExcelColumnBean.java
 - src/main/java/org/embulk/parser/poi_excel/bean/PoiExcelColumnIndex.java
 - src/main/java/org/embulk/parser/poi_excel/bean/PoiExcelSheetBean.java
+- src/main/java/org/embulk/parser/poi_excel/bean/util/PoiExcelCellAddress.java
+- src/main/java/org/embulk/parser/poi_excel/bean/util/SearchMergedCell.java
 - src/main/java/org/embulk/parser/poi_excel/visitor/AbstractPoiExcelCellAttributeVisitor.java
 - src/main/java/org/embulk/parser/poi_excel/visitor/PoiExcelCellCommentVisitor.java
 - src/main/java/org/embulk/parser/poi_excel/visitor/PoiExcelCellFontVisitor.java
 - src/main/java/org/embulk/parser/poi_excel/visitor/PoiExcelCellStyleVisitor.java
+- src/main/java/org/embulk/parser/poi_excel/visitor/PoiExcelCellTypeVisitor.java
 - src/main/java/org/embulk/parser/poi_excel/visitor/PoiExcelCellValueVisitor.java
 - src/main/java/org/embulk/parser/poi_excel/visitor/PoiExcelClientAnchorVisitor.java
 - src/main/java/org/embulk/parser/poi_excel/visitor/PoiExcelColorVisitor.java
@@ -76,29 +88,29 @@ files:
 - src/main/java/org/embulk/parser/poi_excel/visitor/embulk/LongCellVisitor.java
 - src/main/java/org/embulk/parser/poi_excel/visitor/embulk/StringCellVisitor.java
 - src/main/java/org/embulk/parser/poi_excel/visitor/embulk/TimestampCellVisitor.java
+- src/main/java/org/embulk/parser/poi_excel/visitor/util/MergedRegionFinder.java
+- src/main/java/org/embulk/parser/poi_excel/visitor/util/MergedRegionList.java
+- src/main/java/org/embulk/parser/poi_excel/visitor/util/MergedRegionMap.java
+- src/main/java/org/embulk/parser/poi_excel/visitor/util/MergedRegionNothing.java
 - src/test/java/org/embulk/parser/EmbulkPluginTester.java
 - src/test/java/org/embulk/parser/EmbulkTestFileInputPlugin.java
 - src/test/java/org/embulk/parser/EmbulkTestOutputPlugin.java
 - src/test/java/org/embulk/parser/EmbulkTestParserConfig.java
 - src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin.java
+- src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_cellAddress.java
 - src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_cellComment.java
 - src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_cellError.java
 - src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_cellFont.java
 - src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_cellStyle.java
+- src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_cellType.java
 - src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_columnNumber.java
 - src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_constant.java
 - src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_convertError.java
+- src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_formula.java
+- src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_mergedCell.java
 - src/test/java/org/embulk/parser/poi_excel/TestPoiExcelParserPlugin_sheets.java
 - src/test/resources/org/embulk/parser/poi_excel/test1.xls
 - src/test/resources/org/embulk/parser/poi_excel/test2.xlsx
-- classpath/commons-codec-1.9.jar
-- classpath/embulk-parser-poi_excel-0.1.4.jar
-- classpath/embulk-standards-0.7.5.jar
-- classpath/poi-3.13.jar
-- classpath/poi-ooxml-3.13.jar
-- classpath/poi-ooxml-schemas-3.13.jar
-- classpath/stax-api-1.0.1.jar
-- classpath/xmlbeans-2.6.0.jar
 homepage: https://github.com/hishidama/embulk-parser-poi_excel
 licenses:
 - MIT
@@ -109,17 +121,17 @@ require_paths:
 - lib
 required_ruby_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 required_rubygems_version: !ruby/object:Gem::Requirement
   requirements:
-  - - '>='
+  - - ">="
     - !ruby/object:Gem::Version
       version: '0'
 requirements: []
 rubyforge_project:
-rubygems_version: 2.1.9
+rubygems_version: 2.6.8
 signing_key:
 specification_version: 4
 summary: Apache POI Excel parser plugin for Embulk