SQL group by and wmsys.wm_concat 的使用和乱码问题

今天做那个自定义报表做了二三天，写了三天的SQL，还是有问题。

问题是：做现在要以wmps_traceID 为主唯一查询，不要出现重复，然后要把价格累加。 还有一个 ws.userdefined5 字段里面有描述，要把 ws.userdefined5这个字段 不同的内容用 , 放一起显示出来，还不能重复。

业务逻辑就是  以箱子id为主，一个箱子里面 有很多个产品，然后每个产品的价格和数量不一样。要把一个箱子里面的所以的产品价格*数量 总价格显示出来，然后还要把每个产品的描述放一起用 , 分开。 

最后解决的办法是：

group by 把需要合并的字段去点 就不会出现重复的了。

wm_concat（逗号分隔，不重复的话wm_concat(distinct ...)）

wm_concat 乱码问题 加个to_char()就可以啦。

example: replace(to_char(wmsys.wm_concat(distinct to_char(ws.userdefined5))),',',',')as fld_edi_description,

select 
        TO_CHAR(SYSDATE,'yyyymmdd') as fld_TM_manifest_id,
        1 as fld_TM_cartons ,
        nvl(ps.wmps_grossweight,0)as fld_TM_cbm ,
        nvl(ps.wmps_cubic,0)as fld_TM_weight ,
        'AIR'as  fld_TM_container_type,
        '' as fld_TM_container_no,
        'AUD' as fld_TM_seal_no,
        ''as fld_TM_vessel_name,
        TO_CHAR(SYSDATE+1,'DD/MM/YYYY') as fld_TM_edt,
        ps.wmps_traceID as fld_TMD_connote_no,
        p.userdefined2 as fld_TMD_order_no,  
        1 as fld_TMD_cartons,
        nvl(ps.wmps_cubic, 0) as fld_TMD_cbm,
        nvl(ps.wmps_grossweight, 0) as fld_TMD_weight,
        to_char(sysdate,'Month dd yyyy hh:miAM','NLS_DATE_LANGUAGE = American')||'_BWB_Manifest' as fld_TO_filename,
        TO_CHAR(SYSDATE+3,'dd/mm/yyyy') as fld_TO_import_date,
        o.wmor_soreference2 as fld_TO_order_no,
        1 as fld_TO_carrier,
        o.wmor_consigneename as fld_TO_surname,
        NVL(o.wmor_c_address1, ' ') as fld_TO_address_1,
        NVL(o.wmor_c_address2, ' ') as fld_TO_address_2,
        o.wmor_c_country as fld_TO_country_code,
        o.wmor_c_city as fld_TO_suburb,
        NVL(o.wmor_c_province, ' ') as fld_TO_state,
        o.wmor_c_zip as fld_TO_postcode,
        NVL(o.wmor_c_tel1, ' ') as fld_TO_telephone,
        'AeParcelPac' as fld_TO_delivery_instr,
        NVL(o.wmor_c_address3, ' ') as fld_TO_address_3,
        NVL(ps.wmps_grossweight, 0) as fld_TO_weight,
        1 as fld_TOL_line_no,
        '' as fld_TOL_product_no,
        '' as fld_TOL_item_no,   
        1 as fld_TOL_quantity,  
        TO_CHAR(SYSDATE+5,'DD/MM/YY') as fld_TOL_promised_date,   
        sum(NVL(os.userdefined2,0)* os.wmos_qtypicked) as fld_TOL_price,
        sum(NVL(os.userdefined2,0)* os.wmos_qtypicked)   as fld_TOL_total_amount, 
        replace(to_char(wmsys.wm_concat(distinct to_char(ws.userdefined5))),',',',')as fld_edi_description, 
        'Jeanswest' as mame,
        '11th Floor, Unit C, 2 Tins Centre'as addr1,
        'Hung Cheung Road 3' as addr2,
        'Tuen Mun' as city,
        'Hong Kong' as state,
        'NA' as postcode,
        'HK'as country


from wm_out_packing_summary ps,wm_out_packing p,wm_out_order o,wm_out_order_sku os,cd_wh_sku ws
where 
     ps.wmps_traceid=p.wmop_traceid
and  p.wmop_orderno=o.wmor_order_no
and  p.wmop_sku_code=os.wmos_sku_code
and  os.wmos_order_no=o.wmor_order_no
and  ws.cdsk_sku_code=os.wmos_sku_code
group by
ps.wmps_traceID,o.wmor_soreference2,o.wmor_consigneename,o.wmor_c_country,o.wmor_c_city,o.wmor_c_zip,ps.wmps_grossweight,
ps.wmps_cubic,ps.wmps_cubic,ps.wmps_grossweight,o.wmor_c_address1,o.wmor_c_address2,o.wmor_c_province,o.wmor_c_tel1,o.wmor_c_address3,ps.wmps_grossweight,p.userdefined2;

  写了二三天SQL，木有解决问题呀。 头疼，最后跑到CSDN论坛里面去问高手很快就解决了。 感激不尽，心底暗下决心 ，要好好学习，多逛逛论坛，尽自己能力帮别人解决问题。