Tinja is a tiny (~200 lines of code) template engine for C++17, loosely inspired by
inja and mustache. It is
"logic-less" because no complex statements within your templates are needed. Instead, there
are only two kind of tags: variables {{name}} and arrays {[Hello {{name}}!]}
which fulfill most needs for anyone.
- Single header only
- Zero dependencies
- Logic less
- Super fast (outperforms any other template engine)
- Highly memory efficient (runs on ESP8266)
#include <tinja.hpp>
tinja::Template templ("Hello {{name}}!");
tinja::Template::Tokens tokens;
tinja::DataMap data;
data["name"] = "world";
templ.renderTo(data, tokens); // Renders "Hello world!"
templ.parse("{[Hello {{others}}!]}");
data["others"] = tinja::Strings { "Mike", "Charly", "Leo" };
tokens = doc.render(data); // Returns "Hello Mike!", "Hello Charly!", "Hello Leo!"If you render an array and have multiple tags provided, it will be rendered as often as the smallest provided array size. If a variable is provided, it will be simply repeated for each loop:
tinja::Template templ("{[{{firstArray}}: {{variable}} {{secondArray}}!]}");
tinja::Template::Tokens tokens;
tinja::DataMap data;
data["variable"] = "Hello";
data["firstArray"] = tinja::Strings { "1", "2", "3", "4" };
data["secondArray"] = tinja::Strings { "Mike", "Charly", "Leo" };
templ.renderTo(data, tokens); // Renders "1: Hello Mike!", "2: Hello Charly!", "3: Hello Leo!"Tinja is a single header library, which can be downloaded directly from the include/ folder.
To evaluate performance there is a benchmark provided. Comparison is done with several other template engines:
There are multiple test scenarios to try out:
- cold: overall time of parse once and render once
- preprocessed: time to render basic template (taking out parsing step)
- arrays: time to render complex template (with loops)
...............................................................................
cold samples iterations mean
-------------------------------------------------------------------------------
micro_mustache 100 1 25.5542 us
kainjow_mustache 100 1 147.675 us
inja 100 1 37.9807 us
bustache --nlohmann --to_string 100 2 13.0128 us
tinja 100 8 3.40686 us
tinja --32_nodes_reserved 100 9 2.83972 us
tinja --to_string 100 7 3.92889 us
tinja --32_nodes_reserved --to_string 100 8 3.34835 us
-------------------------------------------------------------------------------
preparsed samples iterations mean
-------------------------------------------------------------------------------
micro_mustache 100 1 25.9838 us
kainjow_mustache 100 3 10.9406 us
inja 100 5 6.13677 us
bustache --nlohmann --to_string 100 11 2.30991 us
tinja 100 22 1.08674 us
tinja --to_string 100 16 1.57428 us
-------------------------------------------------------------------------------
arrays samples iterations mean
-------------------------------------------------------------------------------
kainjow_mustache 100 1 60.323 us
inja 100 1 111.585 us
bustache --nlohmann --to_string 100 2 23.8236 us
tinja 100 2 16.2217 us
tinja --to_string 100 2 19.8571 us
Lower is better.
Micro Mustache performs quite nice despite its simple implementation but cannot preparse nor loop. Kainjow Mustache shows bad parsing performance but outperforms inja when it comes to looping. Inja shows nice performance only for simple templates. Bustache has the best performance when it comes to a full-featured mustache implementation. Tinja outperforms Bustache by a factor of 1.2 to 4.6 depending on the scenario but does not offer a complete mustache implementation.
Note: scale is log2.
Tinja is Free Software: You can use, study, share and improve it at your will. Specifically you can redistribute and/or modify it under the terms of the [GNU General Public License][3] as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.