如何获取JSON中的数值

Question

我正在做一个我需要使用OCR的项目。将检索上载图像和图像上的文本。 当我正在使用这个OCR时，我得到的响应是JSON格式。我希望以“text”为键获取所有数值。

我只设法获得看起来像这样的文本中的所有值

我希望结果像

1,381
1,210
1,240
1,160
1,090
1,020

这是JSON

   {
      "boundingBox": "1087,1201,228,83",
      "words": [
        {
          "boundingBox": "1087,1201,228,83",
          "text": "Sonic"
        }
      ]
    },
    {
      "boundingBox": "882,1287,700,55",
      "words": [
        {
          "boundingBox": "882,1287,151,49",
          "text": "Bacon"
        },
        {
          "boundingBox": "1048,1287,170,49",
          "text": "Double"
        },
        {
          "boundingBox": "1232,1287,350,55",
          "text": "Cheeseburger"
        }
      ]
    },
    {
      "boundingBox": "1014,1391,418,109",
      "words": [
        {
          "boundingBox": "1014,1391,244,109",
          "text": "1,240"
        },
        {
          "boundingBox": "1287,1393,145,97",
          "text": "cal"
        }
      ]
    },
    {
      "boundingBox": "853,1422,59,59",
      "words": [
        {
          "boundingBox": "853,1422,59,59",
          "text": "O"
        }
      ]
    },
    {
      "boundingBox": "970,1655,478,90",
      "words": [
        {
          "boundingBox": "970,1655,278,90",
          "text": "Burger"
        },
        {
          "boundingBox": "1270,1655,178,90",
          "text": "King"
        }
      ]
    },
    {
      "boundingBox": "917,1753,630,55",
      "words": [
        {
          "boundingBox": "917,1754,142,54",
          "text": "Triple"
        },
        {
          "boundingBox": "1071,1754,225,54",
          "text": "Whooper"
        },
        {
          "boundingBox": "1308,1753,239,50",
          "text": "Sandwich"
        }
      ]
    },
    {
      "boundingBox": "853,1855,576,109",
      "words": [
        {
          "boundingBox": "853,1892,59,59",
          "text": "e"
        },
        {
          "boundingBox": "1018,1855,236,109",
          "text": "1,160"
        },
        {
          "boundingBox": "1283,1857,146,97",
          "text": "cal"
        }
      ]
    }

这是我使用javascript和html的功能

function traverseJSON (jsonStruct, initialValue) {
	
	for (x in jsonStruct) {
		if (typeof(jsonStruct[x]) == "object") {
			initialValue = traverseJSON(jsonStruct[x], initialValue);
		} else {
			if(x == "text"){
				initialValue += jsonStruct[x] + " ";	
			}
			
		}
		
	}
	return initialValue;
	
}

        .done(function(data) {
            // Show formatted JSON on webpage.
			traverseThis(JSON.stringify(data, null, 2));
        })
        
        

Answer 1

如果我理解你的问题：

function getNumbers(data = []) {
  return data.reduce((arr, entry) => {
    const withNumbers = entry.words.filter(wordEntry => !!Number(wordEntry.text.replace(",", ".")))
    return arr.concat(withNumbers);
  }, []);
}

Answer 2

使用jsonpath - documentChanges获取json中的所有文字。

输入的示例输出

$[*].words[*].text

使用正则表达式 - "Sonic", "Bacon", "Double", "Cheeseburger", "1,240", "cal", "O", "Burger", "King", "Triple", "Whooper", "Sandwich", "e", "1,160", "cal" 来解析此输出。您可以在此处here在线试用。

Answer 3

请检查以下代码是否符合您的要求。我刚刚使用纯Javascript函数isNaN()和replace()实现了这一点。

    let input=[{
    	"boundingBox": "1087,1201,228,83",
    	"words": [{
    		"boundingBox": "1087,1201,228,83",
    		"text": "Sonic"
    	}]
    }, {
    	"boundingBox": "882,1287,700,55",
    	"words": [{
    			"boundingBox": "882,1287,151,49",
    			"text": "Bacon"
    		},
    		{
    			"boundingBox": "1048,1287,170,49",
    			"text": "Double"
    		},
    		{
    			"boundingBox": "1232,1287,350,55",
    			"text": "Cheeseburger"
    		}
    	]
    }, {
    	"boundingBox": "1014,1391,418,109",
    	"words": [{
    			"boundingBox": "1014,1391,244,109",
    			"text": "1,240"
    		},
    		{
    			"boundingBox": "1287,1393,145,97",
    			"text": "cal"
    		}
    	]
    }, {
    	"boundingBox": "853,1422,59,59",
    	"words": [{
    		"boundingBox": "853,1422,59,59",
    		"text": "O"
    	}]
    }, {
    	"boundingBox": "970,1655,478,90",
    	"words": [{
    			"boundingBox": "970,1655,278,90",
    			"text": "Burger"
    		},
    		{
    			"boundingBox": "1270,1655,178,90",
    			"text": "King"
    		}
    	]
    }, {
    	"boundingBox": "917,1753,630,55",
    	"words": [{
    			"boundingBox": "917,1754,142,54",
    			"text": "Triple"
    		},
    		{
    			"boundingBox": "1071,1754,225,54",
    			"text": "Whooper"
    		},
    		{
    			"boundingBox": "1308,1753,239,50",
    			"text": "Sandwich"
    		}
    	]
    }, {
    	"boundingBox": "853,1855,576,109",
    	"words": [{
    			"boundingBox": "853,1892,59,59",
    			"text": "e"
    		},
    		{
    			"boundingBox": "1018,1855,236,109",
    			"text": "1,160"
    		},
    		{
    			"boundingBox": "1283,1857,146,97",
    			"text": "cal"
    		}
    	]
    }]

    function traverseJSON (jsonStruct, initialValue) {
    	
    	for (x in jsonStruct) {
    		if (typeof(jsonStruct[x]) == "object") {
    			initialValue = traverseJSON(jsonStruct[x], initialValue);
    		} else {
    			if(x == "text"){
            if(!isNaN(jsonStruct[x].replace(',',''))){
    				initialValue += jsonStruct[x] + "\n";
             }
    			}
    			
    		}
    		
    	}
    	return initialValue;
    	
    }

console.log(traverseJSON(input,''))