How to solve the problem of content loss when PHP uploads graphic messages with multiple pictures to the WeChat public platform?

1. Question:
Upload an article in wordpress as permanent material to WeChat:
When there is only one picture in the article, it can be uploaded normally, and when viewing this material on the WeChat public platform, both text content and pictures can be Displayed in full.
Add another picture to the article (when the article picture is larger than 1), when viewing this material in the WeChat background, it is found that the article is incomplete: all content starting from the first picture is lost.
Additional instructions:
(1) I wrote a plug-in to upload WordPress articles to the WeChat material library.
Main process: curl grabs the pictures in the article and uploads them to the WeChat material library, obtains the URL and then replaces the picture src in the wordpress article accordingly.
Then construct a json string of image and text message and upload it to WeChat.
(2) No styles are added to the wordpress article, only

, tags
2, code:

// 上传图文消息代码
$raw = Curl::curl_post_wx( 'https://api.weixin.qq.com/cgi-bin/material/add_news?access_token='
    . $token, self::json_article( $post_ids, $token ), false );

/**
 * 将文章内容构造成json字符串
 *
 * @param array $post_ids
 * @param string $token
 * @return string json
 */
private function json_article( $post_ids, $token ) {
    $mp_article = '{"articles":[';
    foreach ( $post_ids as $post_id ) {
        $post = get_post( $post_id );
        $post_urls = self::preg_match_src( $post->post_content );
        if ( $post_urls ) {
            // 如果文章中有图片，将文章中的图片上传至微信素材库
            foreach ( $post_urls as $post_url ) {
                if ( strpos( $post_url, 'qpic.cn' ) === false ) {
                    // 如果图片没有上传到微信服务器
                    $res = Curl::curl_post_wx( 'https://api.weixin.qq.com/cgi-bin/material/add_material?'
                        . 'access_token=' . $token . '&type=image', false, Curl::curl_get_img( $post_url ) );
                    //替换文章内容img中的src地址
                    $post->post_content = str_replace( $post_url, $res['url'], $post->post_content );
                }
                $post->post_content = str_replace( '"', '\"', $post->post_content );
            }
        }

        // 获取缩略图并上传
        // 相应的code省略

        // 构造json
        $mp_article .= '{
            "title": "' . $post->post_title . '",
            "content_source_url": "' . get_permalink( $post_id, false ) . '",
            "thumb_media_id": "' . $thumb_res['media_id'] . '",
            "content": "' . htmlspecialchars_decode( $post->post_content ) . '",
            "show_cover_pic": 0,
        ';
        if ( !empty($post->post_excerpt ) ) {
            $mp_article .= '"digest": "' . $post->post_excerpt . '",';
        }
        $mp_article .= '},';
    }

    return rtrim($mp_article, ',') . ']}';
}

3. Error message: WeChat did not return an error message.
4, Screenshot: None.
5, other references:
I saw tilun_099 encountered similar problems on CSDN: http://bbs.csdn.net/topics/39....
tilun_099's question is: "When uploading multi-image and text materials, if the content of the content is plain text, the upload can be successful, the media_id can be obtained, and then through the group sending interface, the mobile phone can correctly obtain the multi-image and text information. But if the content in the content When tags containing HTML, such as

, etc., although the media_id can be obtained after uploading, after mass sending, the mobile phone will get a single image and text, and this single image and text has a picture, title, and link. , that is, the content is empty. "
tilun_099's final solution to his problem: "Although no one answered, I finally solved it. If the content contains html tags, it needs to be corrected. Escape the content. If it contains content with a "" sign like style="..", you need to pay more attention.

foreach ($news as &$item){
    foreach ($item as $k=>$v){
        if($k =='content'){
            $item[$k] = urlencode(htmlspecialchars(str_replace("\"","'",$v)));
            }else{
                $item[$k] = urlencode($v);
            }
        }
    }

It is necessary to convert the "" quotes in the html into single quotes first, and then convert them into html entities, so that the tags will not be affected when urlencoding is performed.

$data=array("articles"=>$news);
    $data=json_encode($data);
    $end_data= urldecode($data);
    $end_data= htmlspecialchars_decode($end_data);

Before uploading, urldecode the content and convert the html entities into html tags, and it will be OK.
Tested! ”

The problem he and I encountered is somewhat similar, but not exactly the same. I constructed json directly.

Reply content:

1. Question:
Upload an article in wordpress as permanent material to WeChat:
When there is only one picture in the article, it can be uploaded normally, and when viewing this material on the WeChat public platform, both text content and pictures can be Displayed in full.
Add another picture to the article (when the article picture is larger than 1), when viewing this material in the WeChat background, it is found that the article is incomplete: all content starting from the first picture is lost.
Additional explanation:
(1) I wrote a plug-in to upload wordpress articles to the WeChat material library.
Main process: curl grabs the pictures in the article and uploads them to the WeChat material library, obtains the URL and then replaces the picture src in the wordpress article accordingly.
Then construct the image and text message json string and upload it to WeChat.
(2) No styles are added to the wordpress article, only

, tags
2, code:

// 上传图文消息代码
$raw = Curl::curl_post_wx( 'https://api.weixin.qq.com/cgi-bin/material/add_news?access_token='
    . $token, self::json_article( $post_ids, $token ), false );

/**
 * 将文章内容构造成json字符串
 *
 * @param array $post_ids
 * @param string $token
 * @return string json
 */
private function json_article( $post_ids, $token ) {
    $mp_article = '{"articles":[';
    foreach ( $post_ids as $post_id ) {
        $post = get_post( $post_id );
        $post_urls = self::preg_match_src( $post->post_content );
        if ( $post_urls ) {
            // 如果文章中有图片，将文章中的图片上传至微信素材库
            foreach ( $post_urls as $post_url ) {
                if ( strpos( $post_url, 'qpic.cn' ) === false ) {
                    // 如果图片没有上传到微信服务器
                    $res = Curl::curl_post_wx( 'https://api.weixin.qq.com/cgi-bin/material/add_material?'
                        . 'access_token=' . $token . '&type=image', false, Curl::curl_get_img( $post_url ) );
                    //替换文章内容img中的src地址
                    $post->post_content = str_replace( $post_url, $res['url'], $post->post_content );
                }
                $post->post_content = str_replace( '"', '\"', $post->post_content );
            }
        }

        // 获取缩略图并上传
        // 相应的code省略

        // 构造json
        $mp_article .= '{
            "title": "' . $post->post_title . '",
            "content_source_url": "' . get_permalink( $post_id, false ) . '",
            "thumb_media_id": "' . $thumb_res['media_id'] . '",
            "content": "' . htmlspecialchars_decode( $post->post_content ) . '",
            "show_cover_pic": 0,
        ';
        if ( !empty($post->post_excerpt ) ) {
            $mp_article .= '"digest": "' . $post->post_excerpt . '",';
        }
        $mp_article .= '},';
    }

    return rtrim($mp_article, ',') . ']}';
}

3. Error message: WeChat did not return an error message.
4, Screenshot: None.
5, other references:
I saw tilun_099 encountered similar problems on CSDN: http://bbs.csdn.net/topics/39....
tilun_099's question is: "When uploading multi-image and text materials, if the content of the content is plain text, the upload can be successful, the media_id can be obtained, and then through the group sending interface, the mobile phone can correctly obtain the multi-image and text information. But if the content in the content When tags containing HTML, such as

, etc., although the media_id can be obtained after uploading, after mass sending, the mobile phone will get a single image and text, and this single image and text has a picture, title, and link. , that is, the content is empty. "
tilun_099's final solution to his problem: "Although no one answered, I finally solved it. If the content contains html tags, it needs to be corrected. Escape the content. If it contains content with style=".." and "", you need to pay more attention.

foreach ($news as &$item){
    foreach ($item as $k=>$v){
        if($k =='content'){
            $item[$k] = urlencode(htmlspecialchars(str_replace("\"","'",$v)));
            }else{
                $item[$k] = urlencode($v);
            }
        }
    }

It is necessary to convert the "" quotation marks in html into single quotation marks first, and then convert them into html entities, so that the tags will not be affected when urlencoding is performed.

$data=array("articles"=>$news);
    $data=json_encode($data);
    $end_data= urldecode($data);
    $end_data= htmlspecialchars_decode($end_data);

Before uploading, urldecode the content and convert the html entities into html tags, and it will be OK.
Tested! ”

The problem he and I encountered is somewhat similar, but not exactly the same. I constructed json directly.