2 * Copyright (c) 2017 Cisco Systems, Inc. and others. All rights reserved.
4 * This program and the accompanying materials are made available under the
5 * terms of the Eclipse Public License v1.0 which accompanies this distribution,
6 * and is available at http://www.eclipse.org/legal/epl-v10.html
8 package org.opendaylight.mdsal.binding.javav2.generator.util;
10 import com.google.common.annotations.Beta;
11 import com.google.common.base.CharMatcher;
12 import com.google.common.base.Splitter;
13 import com.google.common.collect.ImmutableSet;
14 import com.google.common.collect.ListMultimap;
15 import java.util.Iterator;
16 import java.util.List;
18 import org.opendaylight.mdsal.binding.javav2.generator.context.ModuleContext;
19 import org.opendaylight.mdsal.binding.javav2.model.api.Enumeration;
20 import org.opendaylight.mdsal.binding.javav2.model.api.Enumeration.Pair;
21 import org.opendaylight.mdsal.binding.javav2.util.BindingMapping;
24 * This util class converts every non-java char in identifier to java char by
25 * its unicode name (<a href=
26 * "http://docs.oracle.com/javase/specs/jls/se8/html/jls-3.html#jls-3.8">JAVA SE
27 * SPECIFICATIONS - Identifiers</a>). There are special types of mapping
28 * non-java chars to original identifiers according to specific
29 * {@linkplain JavaIdentifier java type}:
31 * <li>class, enum, interface</li>
34 * <li>without special separator</li>
35 * <li>the first character of identifier, any other first character of
36 * identifier part mapped by non-Java char name from unicode and char in
37 * identifier behind non-java char name are converting to upper case</li>
41 * <li>example* - ExampleAsterisk</li>
42 * <li>example*example - ExampleAserisksExample</li>
43 * <li>\example - ReverseSolidusExample</li>
44 * <li>1example - DigitOneExample</li>
45 * <li>example1 - Example1</li>
46 * <li>int - IntReservedKeyword</li>
47 * <li>con - ConReservedKeyword</li>
52 * <li>enum value, constant</li>
55 * <li>used underscore as special separator</li>
56 * <li>converted identifier to upper case</li>
60 * <li>example* - EXAMPLE_ASTERISK</li>
61 * <li>example*example - EXAMPLE_ASTERISK_EXAMPLE</li>
62 * <li>\example - REVERSE_SOLIDUS_EXAMPLE</li>
63 * <li>1example - DIGIT_ONE_EXAMPLE</li>
64 * <li>example1 - EXAMPLE1</li>
65 * <li>int - INT_RESERVED_KEYWORD</li>
66 * <li>con - CON_RESERVED_KEYWORD</li>
71 * <li>method, variable</li>
75 * <li>without special separator</li>
76 * <li>the first character of identifier is converting to lower case</li>
77 * <li>any other first character of identifier part mapped by non-Java char name
78 * from unicode and char in identifier behind non-java char name are converting
83 * <li>example* - exampleAsterisk</li>
84 * <li>example*example - exampleAserisksExample</li>
85 * <li>\example - reverseSolidusExample</li>
86 * <li>1example - digitOneExample</li>
87 * <li>example1 - example1</li>
88 * <li>int - intReservedKeyword</li>
89 * <li>con - conReservedKeyword</li>
94 * <li>package - full package name (<a href=
95 * "https://docs.oracle.com/javase/tutorial/java/package/namingpkgs.html">
96 * Naming a package</a>)</li>
100 * <li>parts of package name are separated by dots</li>
101 * <li>parts of package name are converting to lower case</li>
102 * <li>if parts of package name are reserved Java or Windows keywords, such as
103 * 'int' the suggested convention is to add an underscore to keyword</li>
104 * <li>dash is parsed as underscore according to <a href=
105 * "https://docs.oracle.com/javase/tutorial/java/package/namingpkgs.html">
106 * Naming a package</a></li>
110 * <li>org.example* - org.exampleasterisk</li>
111 * <li>org.example*example - org.exampleasteriskexample</li>
112 * <li>org.\example - org.reversesolidusexample</li>
113 * <li>org.1example - org.digitoneexample</li>
114 * <li>org.example1 - org.example1</li>
115 * <li>org.int - org.int_</li>
116 * <li>org.con - org.con_</li>
117 * <li>org.foo-cont - org.foo_cont</li>
124 * There is special case in CLASS, INTERFACE, ENUM, ENUM VALUE, CONSTANT, METHOD
125 * and VARIABLE if identifier contains single dash - then the converter ignores
126 * the single dash in the way of the non-java chars. In other way, if dash is
127 * the first or the last char in the identifier or there is more dashes in a row
128 * in the identifier, then these dashes are converted as non-java chars.
131 * <li>class, enum, interface</li>
134 * <li>foo-cont - FooCont</li>
135 * <li>foo--cont - FooHyphenMinusHyphenMinusCont</li>
136 * <li>-foo - HyphenMinusFoo</li>
137 * <li>foo- - FooHyphenMinus</li>
140 * <li>enum value, constant
143 * <li>foo-cont - FOO_CONT</li>
144 * <li>foo--cont - FOO_HYPHEN_MINUS_HYPHEN_MINUS_CONT</li>
145 * <li>-foo - HYPHEN_MINUS_FOO</li>
146 * <li>foo- - FOO_HYPHEN_MINUS</li>
149 * <li>method, variable</li>
152 * <li>foo-cont - fooCont</li>
153 * <li>foo--cont - fooHyphenMinusHyphenMinusCont</li>
154 * <li>-foo - hyphenMinusFoo</li>
155 * <li>foo- - fooHyphenMinus</li>
160 * Next special case talks about normalizing class name which already exists in
161 * package - but with different camel cases (foo, Foo, fOo, ...). To every next
162 * classes with same names will by added their actual rank (serial number),
163 * except the first one. This working for CLASS, ENUM and INTEFACE java
164 * identifiers. If there exist the same ENUM VALUES in ENUM (with different
165 * camel cases), then it's parsed with same logic like CLASSES, ENUMS and
166 * INTERFACES but according to list of pairs of their ENUM parent. Example:
169 * <li>class, enum, interface</li>
172 * <li>package name org.example, class (or interface or enum) Foo - normalized
174 * <li>package name org.example, class (or interface or enum) fOo - normalized
178 * <li>enum value</li>
191 * <li>YANG enum values will be mapped to 'FOO' and 'FOO_1' Java enum
198 public final class JavaIdentifierNormalizer {
200 public static final Set<String> SPECIAL_RESERVED_PATHS = ImmutableSet.of(
201 "org.opendaylight.yangtools.concepts",
202 "org.opendaylight.yangtools.yang.common",
203 "org.opendaylight.yangtools.yang.model",
204 "org.opendaylight.mdsal.binding.javav2.spec",
208 private static final char UNDERSCORE = '_';
209 private static final char DASH = '-';
210 private static final String RESERVED_KEYWORD = "reserved_keyword";
211 private static final Set<String> PRIMITIVE_TYPES = ImmutableSet.of("char[]", "byte[]");
213 private static final CharMatcher DASH_MATCHER = CharMatcher.is(DASH);
214 private static final CharMatcher DASH_OR_SPACE_MATCHER = CharMatcher.anyOf(" -");
215 private static final CharMatcher UNDERSCORE_MATCHER = CharMatcher.is(UNDERSCORE);
216 private static final Splitter DOT_SPLITTER = Splitter.on('.');
217 private static final Splitter UNDERSCORE_SPLITTER = Splitter.on(UNDERSCORE);
219 // Converted to lower case
220 private static final Set<String> WINDOWS_RESERVED_WORDS = BindingMapping.WINDOWS_RESERVED_WORDS.stream()
221 .map(String::toLowerCase).collect(ImmutableSet.toImmutableSet());
223 private JavaIdentifierNormalizer() {
224 throw new UnsupportedOperationException("Util class");
229 * According to <a href="https://tools.ietf.org/html/rfc7950#section-9.6.4">YANG RFC 7950</a>,
230 * all assigned names in an enumeration MUST be unique. Created names are contained in the list
231 * of {@link Enumeration.Pair}. This method adds actual index with underscore behind name of new
232 * enum value only if this name already exists in one of the list of {@link Enumeration.Pair}.
233 * Then, the name will be converted to java chars according to {@link JavaIdentifier#ENUM_VALUE}
245 * YANG enum values will be mapped to 'FOO' and 'FOO_1' Java enum values.
248 * - name of new enum value
250 * - list of all actual enum values
251 * @return converted and fixed name of new enum value
253 public static String normalizeEnumValueIdentifier(final String name, final List<Pair> values) {
254 return convertIdentifierEnumValue(name, name, values, 1);
258 * Normalizing full package name by non java chars and reserved keywords.
260 * @param fullPackageName
261 * - full package name
262 * @return normalized name
264 public static String normalizeFullPackageName(final String fullPackageName) {
265 final Iterator<String> it = DOT_SPLITTER.split(fullPackageName).iterator();
267 return fullPackageName;
270 final StringBuilder sb = new StringBuilder(fullPackageName.length());
272 sb.append(normalizePartialPackageName(it.next()));
274 return sb.toString();
281 * Normalizing part of package name by non java chars.
283 * @param packageNamePart
284 * - part of package name
285 * @return normalized name
287 static String normalizePartialPackageName(final String packageNamePart) {
288 // if part of package name consist from java or windows reserved word, return it with
289 // underscore at the end and in lower case
290 final String lowerPart = packageNamePart.toLowerCase();
291 if (BindingMapping.JAVA_RESERVED_WORDS.contains(lowerPart) || WINDOWS_RESERVED_WORDS.contains(lowerPart)) {
292 return lowerPart + UNDERSCORE;
295 final String normalizedPart = DASH_MATCHER.replaceFrom(packageNamePart, UNDERSCORE);
297 final StringBuilder sb = new StringBuilder();
298 final StringBuilder innerSb = new StringBuilder();
299 for (int i = 0; i < normalizedPart.length(); i++) {
300 final char c = normalizedPart.charAt(i);
301 if (c == UNDERSCORE) {
302 if (innerSb.length() != 0) {
303 sb.append(normalizeSpecificIdentifier(innerSb.toString(), JavaIdentifier.PACKAGE));
304 innerSb.setLength(0);
306 sb.append(UNDERSCORE);
311 if (innerSb.length() != 0) {
312 sb.append(normalizeSpecificIdentifier(innerSb.toString(), JavaIdentifier.PACKAGE));
314 // returned normalized part of package name
315 return sb.toString();
319 * Find and convert non Java chars in identifiers of generated transfer objects, initially
320 * derived from corresponding YANG according to
321 * <a href="http://docs.oracle.com/javase/specs/jls/se8/html/jls-3.html#jls-3.8"> Java
322 * Specifications - Identifiers</a>. If there is more same class names at the same package, then
323 * append rank (serial number) to the end of them. Works for class, enum, interface.
326 * - package of identifier
328 * - name of identifier
329 * @return - java acceptable identifier
331 static String normalizeClassIdentifier(final String packageName, final String className,
332 ModuleContext context) {
333 if (packageName.isEmpty() && PRIMITIVE_TYPES.contains(className)) {
336 for (final String reservedPath : SPECIAL_RESERVED_PATHS) {
337 if (packageName.startsWith(reservedPath)) {
341 final String convertedClassName = normalizeSpecificIdentifier(className, JavaIdentifier.CLASS);
343 // if packageName contains class name at the end, then the className is name of inner class
344 final String basePackageName;
345 final int lastDot = packageName.lastIndexOf('.');
346 if (lastDot != -1 && Character.isUpperCase(packageName.charAt(lastDot + 1))) {
347 // ignore class name in package name - inner class name has to be normalized according to original package
349 basePackageName = packageName.substring(0, lastDot);
351 basePackageName = packageName;
354 return normalizeClassIdentifier(basePackageName, convertedClassName, convertedClassName, 1, context);
358 * Find and convert non Java chars in identifiers of generated transfer objects, initially
359 * derived from corresponding YANG.
362 * - name of identifier
363 * @param javaIdentifier
364 * - java type of identifier
365 * @return - java acceptable identifier
367 public static String normalizeSpecificIdentifier(final String identifier, final JavaIdentifier javaIdentifier) {
368 // if identifier isn't PACKAGE type then check it by reserved keywords
369 if (javaIdentifier != JavaIdentifier.PACKAGE) {
370 final String lower = identifier.toLowerCase();
371 if (BindingMapping.JAVA_RESERVED_WORDS.contains(lower) || WINDOWS_RESERVED_WORDS.contains(lower)) {
372 return fixCasesByJavaType(lower + UNDERSCORE + RESERVED_KEYWORD, javaIdentifier);
376 // check and convert first char in identifier if there is non-java char
377 final StringBuilder sb = new StringBuilder();
378 final char firstChar = identifier.charAt(0);
379 if (!Character.isJavaIdentifierStart(firstChar)) {
380 // converting first char of identifier
381 sb.append(convertFirst(firstChar, existNext(identifier, 0)));
383 sb.append(firstChar);
385 // check and convert other chars in identifier, if there is non-java char
386 for (int i = 1; i < identifier.length(); i++) {
387 final char actualChar = identifier.charAt(i);
388 // ignore single dash as non java char - if there is more dashes in a row or dash is as
389 // the last char in identifier then parse these dashes as non java chars
390 if (actualChar == DASH && existNext(identifier, i)) {
391 if (identifier.charAt(i - 1) != DASH && identifier.charAt(i + 1) != DASH) {
392 sb.append(UNDERSCORE);
396 if (!Character.isJavaIdentifierPart(actualChar)) {
397 // prepare actual string of sb for checking if underscore exist on position of the last char
398 sb.append(convert(actualChar, existNext(identifier, i), sb.charAt(sb.length() - 1)));
400 sb.append(actualChar);
403 // apply camel case in appropriate way
404 return fixCasesByJavaType(sb.toString().replace("__", "_").toLowerCase(), javaIdentifier);
408 * Checking while there doesn't exist any class name with the same name
409 * (regardless of camel cases) in package.
412 * - package of class name
413 * @param origClassName
414 * - original class name
415 * @param actualClassName
416 * - actual class name with rank (serial number)
418 * - actual rank (serial number)
419 * @return converted identifier
421 private static String normalizeClassIdentifier(final String packageName, final String origClassName,
422 final String actualClassName, final int rank, ModuleContext context) {
424 final ListMultimap<String, String> packagesMap = context.getPackagesMap();
426 synchronized (packagesMap) {
427 if (packagesMap.containsKey(packageName)) {
428 for (final String existingName : packagesMap.get(packageName)) {
429 if (actualClassName.equalsIgnoreCase(existingName)) {
430 return normalizeClassIdentifier(packageName, origClassName, origClassName + rank,
435 context.putToPackagesMap(packageName, actualClassName);
436 return actualClassName;
441 * Fix cases of converted identifiers by Java type
443 * @param convertedIdentifier
444 * - converted identifier
445 * @param javaIdentifier
446 * - java type of identifier
447 * @return converted identifier with right cases according to java type
449 private static String fixCasesByJavaType(final String convertedIdentifier, final JavaIdentifier javaIdentifier) {
450 switch (javaIdentifier) {
454 return capitalize(fixCases(convertedIdentifier));
457 return convertedIdentifier.toUpperCase();
460 return fixCases(convertedIdentifier);
462 return UNDERSCORE_MATCHER.removeFrom(convertedIdentifier);
464 throw new IllegalArgumentException("Unknown java type of identifier : " + javaIdentifier.toString());
469 * Delete unnecessary chars in converted identifier and apply camel case in appropriate way.
471 * @param convertedIdentifier
472 * - original converted identifier
473 * @return resolved identifier
475 private static String fixCases(final String convertedIdentifier) {
476 if (convertedIdentifier.indexOf(UNDERSCORE) == -1) {
477 return convertedIdentifier;
480 final StringBuilder sb = new StringBuilder(convertedIdentifier.length());
481 final Iterator<String> it = UNDERSCORE_SPLITTER.split(convertedIdentifier).iterator();
482 sb.append(it.next());
483 while (it.hasNext()) {
484 sb.append(capitalize(it.next()));
486 return sb.toString();
490 * Check if there exist next char in identifier behind actual char position
493 * - original identifier
495 * - actual char position
496 * @return true if there is another char, false otherwise
498 private static boolean existNext(final String identifier, final int actual) {
499 return identifier.length() > actual + 1;
503 * Converting first char of identifier. This happen only if this char is
509 * - existing of next char behind actual char
510 * @return converted char
512 private static String convertFirst(final char c, final boolean existNext) {
513 final String name = DASH_OR_SPACE_MATCHER.replaceFrom(Character.getName(c), UNDERSCORE);
514 return existNext ? name + '_' : name;
518 * Converting any char in java identifier, This happen only if this char is
524 * - existing of next char behind actual char
525 * @param partialLastChar
526 * - last char of partial converted identifier
527 * @return converted char
529 private static String convert(final char c, final boolean existNext, final char partialLastChar) {
530 return partialLastChar == '_' ? convertFirst(c, existNext) : "_" + convertFirst(c, existNext);
534 * Capitalize input string
537 * - string to be capitalized
539 private static String capitalize(final String identifier) {
540 return identifier.substring(0, 1).toUpperCase() + identifier.substring(1);
543 private static String convertIdentifierEnumValue(final String name, final String origName, final List<Pair> values,
545 String newName = name;
546 for (final Pair pair : values) {
547 if (name.equalsIgnoreCase(pair.getName()) || name.equalsIgnoreCase(pair.getMappedName())) {
548 int actualRank = rank;
549 final String actualName = origName + UNDERSCORE + actualRank;
550 newName = convertIdentifierEnumValue(actualName, origName, values, ++actualRank);
553 return normalizeSpecificIdentifier(newName, JavaIdentifier.ENUM_VALUE);