kamusku: the Go library for Kamus Besar Bahasa Indonesia (KBBI)

This module contains HTTP client and command line interface for official
KBBI web.

1 files changed, 98 insertions, 0 deletions

diff --git a/word.go b/word.go
new file mode 100644
index 0000000..8bc0da0
--- /dev/null
+++ b/word.go
@@ -0,0 +1,98 @@
+// Copyright 2020, Shulhan <ms@kilabit.info>. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package kamusku
+
+import (
+	"bytes"
+
+	"github.com/shuLhan/share/lib/net/html"
+)
+
+//
+// Word store the single root word and its definitions.
+//
+type Word struct {
+	Root       string            `json:"dasar,omitempty"` // The root word
+	Definition []*WordDefinition `json:"definisi"`        // The word definition.
+
+	// Message will contains the information when the word is not found or
+	// the word is informal (kata tidak baku).
+	Message string `json:"pesan,omitempty"`
+
+	err error
+}
+
+//
+// Err return an error from retrieving definition.
+//
+func (word *Word) Err() error {
+	return word.err
+}
+
+//
+// parseHTMLEntri parse HTML body from "/entri/<word>" page to find the
+// definition of the word.
+//
+func (word *Word) parseHTMLEntri(in string, htmlBody []byte) (err error) {
+	iter, err := html.Parse(bytes.NewReader(htmlBody))
+	if err != nil {
+		return err
+	}
+
+	for node := iter.Next(); node != nil; node = iter.Next() {
+		if !node.IsElement() {
+			continue
+		}
+
+		switch node.Data {
+		case tagNameHeader2:
+			word.parseRootWord(node)
+
+		case tagNameOrderedList, tagNameUnorderedList:
+			li := node.GetFirstChild()
+			for li != nil {
+				defKata, err := parseWordDefinition(in, li)
+				if err != nil {
+					word.Message = err.Error()
+					err = nil
+					break
+				}
+				if defKata == nil {
+					break
+				}
+				word.Definition = append(word.Definition, defKata)
+				li = li.GetNextSibling()
+			}
+			next := node.GetNextSibling()
+			iter.SetNext(next)
+		}
+	}
+
+	return nil
+}
+
+//
+// parseRootWord given an HMTL element "h2" find a possible root word and
+// return true; otherwise it will return false.
+//
+func (word *Word) parseRootWord(h2 *html.Node) bool {
+	el := h2.GetFirstChild()
+	if el.Data != tagNameSpan {
+		return false
+	}
+	v := el.GetAttrValue(attrNameClass)
+	if v != attrValueRootWord {
+		return false
+	}
+
+	el = el.GetFirstChild()
+	if el.Data != tagNameAnchor {
+		return false
+	}
+	el = el.GetFirstChild()
+	word.Root = el.Data
+
+	return true
+}