add daochai parser

1 files changed, 187 insertions, 0 deletions

diff --git a/parser/daochai.fnl b/parser/daochai.fnl
new file mode 100644
index 0000000..50ec508
--- /dev/null
+++ b/parser/daochai.fnl
@@ -0,0 +1,187 @@
+(import-macros {: reduce} :lib.macro)
+
+(local peg
+ (if (pick-values 1 (pcall require :lpeg))
+   (require :lpeg)
+   (require :lpeglj)))
+(local parser (require :parser.parser))
+(local number (require :lib.number))
+(local fetcher (require :fetcher))
+
+(fn html-cyrillic->utf [str]
+  (local replacement-map
+   {"А" "А"
+    "Б" "Б"
+    "В" "В"
+    "Г" "Г"
+    "Д" "Д"
+    "Е" "Е"
+    "Ж" "Ж"
+    "З" "З"
+    "И" "И"
+    "Й" "Й"
+    "К" "К"
+    "Л" "Л"
+    "М" "М"
+    "Н" "Н"
+    "О" "О"
+    "П" "П"
+    "Р" "Р"
+    "С" "С"
+    "Т" "Т"
+    "У" "У"
+    "Ф" "Ф"
+    "Х" "Х"
+    "Ц" "Ц"
+    "Ч" "Ч"
+    "Ш" "Ш"
+    "Щ" "Щ"
+    "Ъ" "Ъ"
+    "Ы" "Ы"
+    "Ь" "Ь"
+    "Э" "Э"
+    "Ю" "Ю"
+    "Я" "Я"
+    "а" "а"
+    "б" "б"
+    "в" "в"
+    "г" "г"
+    "д" "д"
+    "е" "е"
+    "ж" "ж"
+    "з" "з"
+    "и" "и"
+    "й" "й"
+    "к" "к"
+    "л" "л"
+    "м" "м"
+    "н" "н"
+    "о" "о"
+    "п" "п"
+    "р" "р"
+    "с" "с"
+    "т" "т"
+    "у" "у"
+    "ф" "ф"
+    "х" "х"
+    "ц" "ц"
+    "ч" "ч"
+    "ш" "ш"
+    "щ" "щ"
+    "ъ" "ъ"
+    "ы" "ы"
+    "ь" "ь"
+    "э" "э"
+    "ю" "ю"
+    "я" "я"})
+
+  (var result str)
+  (each [code letter (pairs replacement-map)]
+    (set result (: result :gsub code letter)))
+  result)
+
+(fn format-url [path page]
+  (.. "https://daochai.ru/" path
+      "/" (if (< 1 page) (.. "page-" page "/") "")))
+
+(local product-peg
+  (*
+    (parser.anywhere
+      (+
+       ;; eager and lazy loaded versions of img
+       (parser.tag :img {:class "ty-pict   cm-image" ;; FRAGILE
+                         :src (peg.Cg (parser.till "\"") :image)
+                         :id "*"
+                         :title "*"
+                         :alt "*"
+                         :srcset "*"
+                         :width "*"
+                         :height "*"})
+       (parser.tag :img {:class "ty-pict   cm-image" ;; FRAGILE
+                         :src "*"
+                         :data-src (peg.Cg (parser.till "\"") :image)
+                         :id "*"
+                         :title "*"
+                         :alt "*"
+                         :data-srcset "*"
+                         :width "*"
+                         :height "*"})))
+    (parser.anywhere
+      (parser.tag :a {:class "product-title"
+                      :href (peg.Cg (parser.till "\"") :url)
+                      :title "*"}
+        (peg.Cg (parser.till "</a>") :title)))
+    (+
+     (*
+      (parser.anywhere
+        (parser.tag :span {:class "ty-price-num" :id "*"}
+          (peg.Cg (parser.till "</span>") :price)))
+      (parser.anywhere
+        ;; "за" and "гр" words are html-encoded for some reason
+        (* "&#1079;&#1072; " (peg.Cg parser.pegs.number :weight) "  &#1075;&#1088;.")))
+     (parser.anywhere
+      (parser.tag :span {:class "ty-price-num" :id "*"}
+       (peg.Cg (parser.till "</span>") :price))))
+    (parser.anywhere
+     ;; "Купить"
+     (parser.tag :span {} "&#1050;&#1091;&#1087;&#1080;&#1090;&#1100;"))))
+
+(fn normalize [product]
+  (local title (html-cyrillic->utf product.title))
+  (local year (parser.guess-year title))
+  (local weight (number.string->number product.weight))
+  (local price (number.string->number product.price))
+
+  {:site "daochai"
+   :title title
+   :url product.url
+   :description nil
+   :image product.image
+   :year year
+   :price price
+   :weight weight
+   :volume (parser.guess-volume title)
+   :price-per (if (and price weight (< 0 weight))
+                (/ (math.ceil (* (/ price weight) 10)) 10)
+                nil)})
+
+(fn products []
+  (fetcher.from-html
+    [{:path "vid-chaya/pu-erh/shu-puer" :tags ["Шу пуэр"]}
+     {:path "vid-chaya/pu-erh/shen" :tags ["Шен пуэр"]}
+     {:path "vid-chaya/ulun" :tags ["Улун"]}
+     {:path "vid-chaya/ulun/fudzjanskie-uluny" :tags ["Улун" "Фудзянь"]}
+     {:path "vid-chaya/ulun/guandunskie-uluny" :tags ["Улун" "Гуандун"]}
+     {:path "vid-chaya/ulun/uishanskie-uluny" :tags ["Улун" "Уишань"]}
+     {:path "vid-chaya/ulun/taiwan-ulun" :tags ["Улун" "Тайвань"]}
+     {:path "vid-chaya/ulun/yunnanskiy-uluny" :tags ["Улун" "Юннань"]}
+     {:path "vid-chaya/krasnyj-chaj" :tags ["Красный чай"]}
+     {:path "vid-chaya/zeljonyj-chaj" :tags ["Зеленый чай"]}
+     {:path "vid-chaya/white" :tags ["Белый чай"]}
+     {:path "vid-chaya/zheltyy-chay" :tags ["Желтый чай"]}
+     {:path "vid-chaya/heicha" :tags ["Хэй ча"]}
+     {:path "posuda/jianshuizitao" :tags ["Посуда" "Чайник"]}
+     {:path "posuda/nisintao" :tags ["Посуда" "Чайник"]}
+     {:path "posuda/chahu-taozi" :tags ["Посуда" "Чайник"]}
+     {:path "posuda/chayniki-iz-chaochzhou" :tags ["Посуда" "Чайник"]}
+     {:path "posuda/jingdezhen" :tags ["Посуда"]}
+     {:path "posuda/chahai" :tags ["Посуда"]}
+     {:path "posuda/gajvan" :tags ["Посуда"]}
+     {:path "posuda/chahaj" :tags ["Посуда"]}
+     {:path "posuda/chaban" :tags ["Посуда"]}
+     {:path "posuda/chajnye-prudy" :tags ["Посуда"]}
+     {:path "posuda/sito" :tags ["Посуда"]}
+     {:path "posuda/posuda-chajnoj-ceremonii" :tags ["Посуда"]}
+     {:path "posuda/termosy" :tags ["Посуда"]}
+     {:path "posuda/alternativa" :tags ["Посуда"]}
+     {:path "tea-accessorize" :tags ["Посуда"]}
+     {:path "chay-i-chan/aroma" :tags ["Благовония"]}
+     {:path "chay-i-chan/kurilnicy-i-podstavki-pod-blagovoniya" :tags ["Благовония"]}
+     {:path "chay-i-chan/chetki" :tags ["Четки"]}
+     {:path "chay-i-chan/dekorirovanie-prostranstva" :tags ["Декор"]}
+     {:path "chay-i-chan/figurki-iz-dereva" :tags ["Фигурки"]}]
+    format-url
+    product-peg
+    normalize))
+
+{:products products :title "DaoChai" :url "https://daochai.ru"}