blob: ba4c668ffb4b63c24cba62ab8f484aebadda593b [file] [log] [blame]
/*
* Copyright (C) 2010 Google Inc.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package com.google.i18n.addressinput.common;
import com.google.i18n.addressinput.common.AddressField.WidthType;
import com.google.i18n.addressinput.common.LookupKey.ScriptType;
import org.json.JSONException;
import org.json.JSONObject;
import org.json.JSONTokener;
import java.util.ArrayList;
import java.util.Collections;
import java.util.EnumSet;
import java.util.List;
import java.util.Map;
import java.util.Set;
/**
* Address format interpreter. A utility to find address format related info.
*/
public final class FormatInterpreter {
private static final String NEW_LINE = "%n";
private final FormOptions.Snapshot formOptions;
/**
* Creates a new instance of {@link FormatInterpreter}.
*/
public FormatInterpreter(FormOptions.Snapshot options) {
Util.checkNotNull(
RegionDataConstants.getCountryFormatMap(), "null country name map not allowed");
Util.checkNotNull(options);
this.formOptions = options;
Util.checkNotNull(getJsonValue("ZZ", AddressDataKey.FMT),
"Could not obtain a default address field order.");
}
/**
* Returns a list of address fields based on the format of {@code regionCode}. Script type is
* needed because some countries uses different address formats for local/Latin scripts.
*
* @param scriptType if {@link ScriptType#LOCAL}, use local format; else use Latin format.
*/
// TODO: Consider not re-doing this work every time the widget is re-rendered.
@SuppressWarnings("deprecation") // For legacy address fields.
public List<AddressField> getAddressFieldOrder(ScriptType scriptType, String regionCode) {
Util.checkNotNull(scriptType);
Util.checkNotNull(regionCode);
String formatString = getFormatString(scriptType, regionCode);
return getAddressFieldOrder(formatString, regionCode);
}
List<AddressField> getAddressFieldOrder(String formatString, String regionCode) {
EnumSet<AddressField> visibleFields = EnumSet.noneOf(AddressField.class);
List<AddressField> fieldOrder = new ArrayList<AddressField>();
// TODO: Change this to just enumerate the address fields directly.
for (String substring : getFormatSubstrings(formatString)) {
// Skips un-escaped characters and new lines.
if (!substring.matches("%.") || substring.equals(NEW_LINE)) {
continue;
}
AddressField field = getFieldForFormatSubstring(substring);
// Accept only the first instance for any duplicate fields (which can occur because the
// string we start with defines format order, which can contain duplicate fields).
if (!visibleFields.contains(field)) {
visibleFields.add(field);
fieldOrder.add(field);
}
}
applyFieldOrderOverrides(regionCode, fieldOrder);
// Uses two address lines instead of street address.
for (int n = 0; n < fieldOrder.size(); n++) {
if (fieldOrder.get(n) == AddressField.STREET_ADDRESS) {
fieldOrder.set(n, AddressField.ADDRESS_LINE_1);
fieldOrder.add(n + 1, AddressField.ADDRESS_LINE_2);
break;
}
}
return Collections.unmodifiableList(fieldOrder);
}
/**
* Returns true if this format substring (e.g. %C) represents an address field. Returns false if
* it is a literal or newline.
*/
private static boolean formatSubstringRepresentsField(String formatSubstring) {
return !formatSubstring.equals(NEW_LINE) && formatSubstring.startsWith("%");
}
/**
* Gets data from the address represented by a format substring such as %C. Will throw an
* exception if no field can be found.
*/
private static AddressField getFieldForFormatSubstring(String formatSubstring) {
return AddressField.of(formatSubstring.charAt(1));
}
/**
* Returns true if the address has any data for this address field.
*/
private static boolean addressHasValueForField(AddressData address, AddressField field) {
if (field == AddressField.STREET_ADDRESS) {
return address.getAddressLines().size() > 0;
} else {
String value = address.getFieldValue(field);
return (value != null && !value.isEmpty());
}
}
private void applyFieldOrderOverrides(String regionCode, List<AddressField> fieldOrder) {
List<AddressField> customFieldOrder = formOptions.getCustomFieldOrder(regionCode);
if (customFieldOrder == null) {
return;
}
// We can assert that fieldOrder and customFieldOrder contain no duplicates.
// We know this by the construction above and in FormOptions but we still have to think
// about fields in the custom ordering which aren't visible (the loop below will fail if
// a non-visible field appears in the custom ordering). However in that case it's safe to
// just ignore the extraneous field.
Set<AddressField> nonVisibleCustomFields = EnumSet.copyOf(customFieldOrder);
nonVisibleCustomFields.removeAll(fieldOrder);
if (nonVisibleCustomFields.size() > 0) {
// Local mutable copy to remove non visible fields - this shouldn't happen often.
customFieldOrder = new ArrayList<AddressField>(customFieldOrder);
customFieldOrder.removeAll(nonVisibleCustomFields);
}
// It is vital for this loop to work correctly that every element in customFieldOrder
// appears in fieldOrder exactly once.
for (int fieldIdx = 0, customIdx = 0; fieldIdx < fieldOrder.size(); fieldIdx++) {
if (customFieldOrder.contains(fieldOrder.get(fieldIdx))) {
fieldOrder.set(fieldIdx, customFieldOrder.get(customIdx++));
}
}
}
/**
* Returns the fields that are required to be filled in for this country. This is based upon the
* "required" field in RegionDataConstants for {@code regionCode}, and handles falling back to
* the default data if necessary.
*/
static Set<AddressField> getRequiredFields(String regionCode) {
Util.checkNotNull(regionCode);
String requireString = getRequiredString(regionCode);
return getRequiredFields(requireString, regionCode);
}
static Set<AddressField> getRequiredFields(String requireString, String regionCode) {
EnumSet<AddressField> required = EnumSet.of(AddressField.COUNTRY);
for (char c : requireString.toCharArray()) {
required.add(AddressField.of(c));
}
return required;
}
private static String getRequiredString(String regionCode) {
String required = getJsonValue(regionCode, AddressDataKey.REQUIRE);
if (required == null) {
required = getJsonValue("ZZ", AddressDataKey.REQUIRE);
}
return required;
}
/**
* Returns the field width override for the specified country, or null if there's none. This is
* based upon the "width_overrides" field in RegionDataConstants for {@code regionCode}.
*/
static WidthType getWidthOverride(AddressField field, String regionCode) {
return getWidthOverride(field, regionCode, RegionDataConstants.getCountryFormatMap());
}
/**
* Visible for Testing - same as {@link #getWidthOverride(AddressField, String)} but testable with
* fake data.
*/
static WidthType getWidthOverride(
AddressField field, String regionCode, Map<String, String> regionDataMap) {
Util.checkNotNull(regionCode);
String overridesString =
getJsonValue(regionCode, AddressDataKey.WIDTH_OVERRIDES, regionDataMap);
if (overridesString == null || overridesString.isEmpty()) {
return null;
}
// The field width overrides string starts with a %, so we skip the first one.
// Example string: "%C:L%S:S" which is a repeated string of
// '<%> field_character <:> width_character'.
for (int pos = 0; pos != -1;) {
int keyStartIndex = pos + 1;
int valueStartIndex = overridesString.indexOf(':', keyStartIndex + 1) + 1;
if (valueStartIndex == 0 || valueStartIndex == overridesString.length()) {
// Malformed string -- % not followed by ':' or trailing ':'
return null;
}
// Prepare for next iteration.
pos = overridesString.indexOf('%', valueStartIndex + 1);
if (valueStartIndex != keyStartIndex + 2 ||
overridesString.charAt(keyStartIndex) != field.getChar()) {
// Key is not a high level field (unhandled by this code) or does not match.
// Also catches malformed string where key is of zero length (skip, not error).
continue;
}
int valueLength = (pos != -1 ? pos : overridesString.length()) - valueStartIndex;
if (valueLength != 1) {
// Malformed string -- value has length other than 1
return null;
}
return WidthType.of(overridesString.charAt(valueStartIndex));
}
return null;
}
/**
* Gets formatted address. For example,
*
* <p> John Doe</br>
* Dnar Corp</br>
* 5th St</br>
* Santa Monica CA 90123 </p>
*
* This method does not validate addresses. Also, it will "normalize" the result strings by
* removing redundant spaces and empty lines.
*/
public List<String> getEnvelopeAddress(AddressData address) {
Util.checkNotNull(address, "null input address not allowed");
String regionCode = address.getPostalCountry();
String lc = address.getLanguageCode();
ScriptType scriptType = ScriptType.LOCAL;
if (lc != null) {
scriptType = Util.isExplicitLatinScript(lc) ? ScriptType.LATIN : ScriptType.LOCAL;
}
List<String> prunedFormat = new ArrayList<String>();
String formatString = getFormatString(scriptType, regionCode);
List<String> formatSubstrings = getFormatSubstrings(formatString);
for (int i = 0; i < formatSubstrings.size(); i++) {
String formatSubstring = formatSubstrings.get(i);
// Always keep the newlines.
if (formatSubstring.equals(NEW_LINE)) {
prunedFormat.add(NEW_LINE);
} else if (formatSubstringRepresentsField(formatSubstring)) {
// Always keep the non-empty address fields.
if (addressHasValueForField(address, getFieldForFormatSubstring(formatSubstring))) {
prunedFormat.add(formatSubstring);
}
} else if (
// Only keep literals that satisfy these 2 conditions:
// (1) Not preceding an empty field.
(i == formatSubstrings.size() - 1 || formatSubstrings.get(i + 1).equals(NEW_LINE)
|| addressHasValueForField(address, getFieldForFormatSubstring(
formatSubstrings.get(i + 1))))
// (2) Not following a removed field.
&& (i == 0 || !formatSubstringRepresentsField(formatSubstrings.get(i - 1))
|| (!prunedFormat.isEmpty()
&& formatSubstringRepresentsField(prunedFormat.get(prunedFormat.size() - 1))))) {
prunedFormat.add(formatSubstring);
}
}
List<String> lines = new ArrayList<>();
StringBuilder currentLine = new StringBuilder();
for (String formatSubstring : prunedFormat) {
if (formatSubstring.equals(NEW_LINE)) {
if (currentLine.length() > 0) {
lines.add(currentLine.toString());
currentLine.setLength(0);
}
} else if (formatSubstringRepresentsField(formatSubstring)) {
switch (getFieldForFormatSubstring(formatSubstring)) {
case STREET_ADDRESS:
// The field "street address" represents the street address lines of an address, so
// there can be multiple values.
List<String> addressLines = address.getAddressLines();
if (addressLines.size() > 0) {
currentLine.append(addressLines.get(0));
if (addressLines.size() > 1) {
lines.add(currentLine.toString());
currentLine.setLength(0);
lines.addAll(addressLines.subList(1, addressLines.size()));
}
}
break;
case COUNTRY:
// Country name is treated separately.
break;
case ADMIN_AREA:
currentLine.append(address.getAdministrativeArea());
break;
case LOCALITY:
currentLine.append(address.getLocality());
break;
case DEPENDENT_LOCALITY:
currentLine.append(address.getDependentLocality());
break;
case RECIPIENT:
currentLine.append(address.getRecipient());
break;
case ORGANIZATION:
currentLine.append(address.getOrganization());
break;
case POSTAL_CODE:
currentLine.append(address.getPostalCode());
break;
case SORTING_CODE:
currentLine.append(address.getSortingCode());
break;
default:
break;
}
} else {
// Not a symbol we recognise, so must be a literal. We append it unchanged.
currentLine.append(formatSubstring);
}
}
if (currentLine.length() > 0) {
lines.add(currentLine.toString());
}
return lines;
}
/**
* Tokenizes the format string and returns the token string list. "%" is treated as an escape
* character. For example, "%n%a%nxyz" will be split into "%n", "%a", "%n", "xyz".
* Escaped tokens correspond to either new line or address fields. The output of this method
* may contain duplicates.
*/
// TODO: Create a common method which does field parsing in one place (there are about 4 other
// places in this library where format strings are parsed).
private List<String> getFormatSubstrings(String formatString) {
List<String> parts = new ArrayList<String>();
boolean escaped = false;
StringBuilder currentLiteral = new StringBuilder();
for (char c : formatString.toCharArray()) {
if (escaped) {
escaped = false;
parts.add("%" + c);
} else if (c == '%') {
if (currentLiteral.length() > 0) {
parts.add(currentLiteral.toString());
currentLiteral.setLength(0);
}
escaped = true;
} else {
currentLiteral.append(c);
}
}
if (currentLiteral.length() > 0) {
parts.add(currentLiteral.toString());
}
return parts;
}
private static String getFormatString(ScriptType scriptType, String regionCode) {
String format = (scriptType == ScriptType.LOCAL)
? getJsonValue(regionCode, AddressDataKey.FMT)
: getJsonValue(regionCode, AddressDataKey.LFMT);
if (format == null) {
format = getJsonValue("ZZ", AddressDataKey.FMT);
}
return format;
}
private static String getJsonValue(String regionCode, AddressDataKey key) {
return getJsonValue(regionCode, key, RegionDataConstants.getCountryFormatMap());
}
/**
* Visible for testing only.
*/
static String getJsonValue(
String regionCode, AddressDataKey key, Map<String, String> regionDataMap) {
Util.checkNotNull(regionCode);
String jsonString = regionDataMap.get(regionCode);
Util.checkNotNull(jsonString, "no json data for region code " + regionCode);
try {
JSONObject jsonObj = new JSONObject(new JSONTokener(jsonString));
if (!jsonObj.has(Util.toLowerCaseLocaleIndependent(key.name()))) {
// Key not found. Return null.
return null;
}
// Gets the string for this key.
String parsedJsonString = jsonObj.getString(Util.toLowerCaseLocaleIndependent(key.name()));
return parsedJsonString;
} catch (JSONException e) {
throw new RuntimeException("Invalid json for region code " + regionCode + ": " + jsonString);
}
}
}