From 3e2a243678512dab4c51c9760c937feb8aa1383f Mon Sep 17 00:00:00 2001
From: zlt <zlt@zlt.com>
Date: Fri, 28 Aug 2026 17:20:58 +0800
Subject: [PATCH] =?UTF-8?q?=E4=BF=AE=E5=A4=8D=E5=85=AC=E7=A4=BA=E5=AF=BC?=
 =?UTF-8?q?=E5=87=BA=EF=BC=9A=E6=A8=A1=E6=9D=BF=E6=AE=8B=E7=95=99=E9=9A=90?=
 =?UTF-8?q?=E8=97=8F=E8=A1=8C=E6=8A=8A=E5=90=88=E8=AE=A1=E8=A1=8C=E4=B8=8E?=
 =?UTF-8?q?=E5=9B=BE=E8=A1=A8=E6=95=B4=E6=AE=B5=E8=97=8F=E6=8E=89=EF=BC=9B?=
 =?UTF-8?q?=E7=BA=A2=E7=BB=BF=E6=9D=A1=E4=BB=B6=E6=A0=BC=E5=BC=8F=E6=8C=89?=
 =?UTF-8?q?=E6=9C=80=E7=BB=88=E8=A1=8C=E4=BD=8D=E9=87=8D=E5=BB=BA=EF=BC=9B?=
 =?UTF-8?q?=E5=87=80=E5=80=BC+=E4=BB=93=E4=BD=8D=E5=8F=8C=E7=B3=BB?=
 =?UTF-8?q?=E5=88=97=E5=B8=A6=E6=A0=87=E8=AE=B0=E5=9B=BE=E8=A1=A8?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 DEVLOG.md                        |  35 +++++++++++++
 app/assets/publish_template.xlsx | Bin 11747 -> 9203 bytes
 app/services/publish_export.py   |  81 ++++++++++++++++++++++++++-----
 scripts/test_wiring.py           |  38 +++++++++++++++
 4 files changed, 141 insertions(+), 13 deletions(-)

diff --git a/DEVLOG.md b/DEVLOG.md
index 7202168..54df35a 100644
--- a/DEVLOG.md
+++ b/DEVLOG.md
@@ -1107,6 +1107,41 @@ init_db 幂等带上)。无新 env。
 
 ---
 
+## 2026-08-28 · 公示导出正式机实锅: 模板隐藏行把合计与图整段藏掉
+
+**做了什么**
+正式机首次导出缺了最底部合计行、图表空框。根因: 公司原表把老平仓行折叠隐藏
+(24~97 行 hidden), 裁模板时 openpyxl 的 delete_rows 不搬 row_dimensions, 这批隐藏
+标记原样进了模板; 表体一长到 24 行 (真机 7 持仓+10 平仓正好踩中), 卖出小计/存量
+合计/底行/图表数据全落隐藏带 —— 此前样张 7+7 只到 23 行, 所以怎么测都完整。
+开发机 openpyxl 3.1.5 实验复现坐实。修法两层: 模板重裁 (行属性只留 1~11 行、隐藏
+标记清零、剔除全部残留条件格式区间与外部工作簿链接, 顺带补回裁剪时丢的「存量合计」
+「初始规模+已回收益后可开仓总金额」两个 B 列标签 —— 与上次「平层卖出」同类伤);
+代码兜底 (进场即清模板行位之外的行属性与残留条件格式, 行高重排时强制 hidden=False)。
+红绿条件格式按最终行位重建: Q/R/S 红涨绿跌字色、持仓 S 列红绿底色 (0 也算红),
+颜色与原表 dxf 逐字节同 (openpyxl 去重后直接命中原 18 个 dxf)。图表按原图重做双系列:
+净值主轴(左)+仓位占比副轴(右), 每点带标记 (首日单点也看得见), plotVisOnly=0,
+smooth=0。O/P 融资两列原表就是折叠的, 属版式, 保留。
+
+**动了哪些文件**
+app/assets/publish_template.xlsx (重裁)、app/services/publish_export.py (卫生+条件
+格式+双系列图)、scripts/test_wiring.py (公示导出用例补大表回归: 7+10 越过隐藏带,
+三行合计/无隐藏行/条件格式区间/双系列带标记断言钉死)。
+
+**部署方式**
+两台都要: git pull 后重新 build 部署 (模板打进镜像, 必须重建)。真机 188 与模拟机
+tlai4090 同法。无新 env、无 DDL。
+
+**真机判收**
+开发机全部 20 套 588 例 ALL PASS (含新回归); 云端 LibreOffice 重算 92 公式 0 错误,
+恒等式核过; 7+10 形状与 30 日净值序列两版渲染逐区目验 (三行合计齐、黄条粉底在、
+双系列图两轴两色带标记)。部署后判收: 重新点「导出公示表」, 底部三行 (卖出小计/
+存量合计/底行) 与图表单点标记都应出现。
+
+**还欠着什么**
+上节点的三个口径拍板照旧欠着; 图表逐像素对齐照旧按需。
+
+---
 <!--
 下一条节点从这里往下写，格式照抄上面：
 ## YYYY-MM-DD · 一句话标题
diff --git a/app/assets/publish_template.xlsx b/app/assets/publish_template.xlsx
index 0c27507ed474ee774256cedbe97b8b30ea1b6d9f..ce861482713577ed6628d4e97e7c39c3bbc31254 100644
GIT binary patch
delta 4468
zcmb7IcQ{;I*B>=9dKZT1Wc1!U(d$JpdW;aGj|d3|i6A5~W)QtZhD7hZ1;LPs5=4!V
zC^34ph<M}P=YG$<@B7!cpJ(si*=3!x_gcTbp0&<0Y+s+s!kC1V2?PRBfQ$@LRNWLt
zh7hOMwR=BN6GFh(+_Ib!o9Z#hu}VEt0tM9X?Q{c6AY%Hu_b0CYx@*jHsQbGFLSyY%
z+8fCNF?D<2k?@4A+m$nTDYunD{JkDf_Q!%HamuISObRp5t)3~r?ttCyMhYdKKP=Y5
zdeq{Go@b(-Tf_ODR}Hi`=-=nN-gZ*w`%sR!N-s7iR8s~38`d`gj_qYeijB0w&up9h
ziLy|v6B-58#Dk;Fp+U%lh`a;6@3GXvW$B5lO$eyUYM78DQfcsK_oAwjX{4Cn_P4kq
zoS}1(N6J_m7BhO8?~Ct1IBFYWw|?F43m}w>^+;Q(&2*G8vNHqt)yc|m0Zd|{kzq`J
zG!rUQpIK6OS{dqx7+-gQ{;8!}I^rlYVk5)aluHIg3>W&MSQ*z*rUZeO^*|st5E?JS
z4zxSk2hS=pM82=I2)n%sFBM(lB@zlu-4^NaQ0Z^GCeK*<wx*4tJ!9IXlbeDrWo``{
zG@J<E)dA);0n_}D_7nMxUrMbKL$8mfjz0$;-r4MUoh$$2Y<e?7i3>j|Wd8MqUWW;G
z*$tj&)_Hi#d_3sxYSLTUh+KFWpbAl|0DL<Fr^K=Cvx81y@{n}@sfuqR!`kT8Z*Jf1
z90CqixoC?4|8!ckzyH)nhbBq3i1x~FlxOc0%zG-o+St#lI+<+NcC60d4_07Le^xvC
zrCLF(v(0JNlLI!J1g}rv*pIZ1%zAhBsq|4^B`$8S0lJ-C8Qzj7gx2uc8F0pZ14`zj
zGLc?af5Emew6y+H>QyGQAAc=FCd}p4rMFZMU_VaIAIwHHTHVRbi>nQ~wYTGjyfF{^
zH0`im@l*PKWXY1=xHE@I2WMI&k;mqjJA0DBz`akI)%Vrj!_$K6CtD+OZP0<-4O+H<
z2Bw~}dXnQA*L?l%jzni#FywYl2(Yr4`#>Vlpf@cH_Sv^D#kYE)ulD=oUTnzGTI#BW
zF(f?JW>`(Xqqn=t+rZ!P?y~!|BFv3t2K^kK(w0b3lOKZ^oGYX3mgyd0d!uV7f`!Yi
zi))cysfKfAh<7|l{#cXp1zb81lRty|HvepE3x~B(emKrRW+Rm<=<XCB1Jr$_^evaI
z=9WIxhR1veYkjNbahS)dyq%AhXJfqUT#{ODfmAI?lPFCKvA5^?fJ>ji?b&$t%+7wQ
z@e_LwRv6`W?GSav-YB>XUKu*-otTtgOPBb>qB2kKkeTn?QMQ1j|Jp5Ykft(tnB|N2
z&2H_v-E`uqd@`A(JQOa$1lYSGLiLTONpGr6@iq1^XGtYa^(X%-@AqRjk>HNlC_L!l
z(2#%i_^_8l{$Om>jD5yb=bpEh1~q3<L60mF^XTdL$mOF%kn8rCXL0<r{R}~s+Nr2u
zNLH32u<c<5OaAE12*vi-*gFFQ;F_7AX9}obtn>LVj@9d|2RHrXJb|mREDp2%jtQc`
z8P>&qat8ad<M0C&oySgQvA{K3RFQ$PH1#I7pk1~u?-9+|h#R}Ikxsw&nl5*WNmdT8
z=PmEreOV;xk@)zCW|6=(VZ%Mgk~u>u#`^+>kL&o!#0;m;{6x=~Js47$u{6$KOgbas
z=a$+n+C#TN@sa&#Bj6$WR>k!US#aUUO@{P&V$L|aLgx?Jm+dufn1JJhpV5fi#O7uU
zfbraqBg_kuF{?33U>=$HrTm^@X_hZz0n9N|SLg(Xo}j5i_daLf9|AOjOUYEMo!-P4
zeNI~tqnc)ziT)sbLAO{sz|#CpAxjSH1dcyAXiy+{sy18z1Z;(VPm0j`DB7$V2oKua
zx7EZHbG}I(`#M=sXzP2|%*8=F`&L~AA;VH>-s{CYrg`P{6V^r5qwICbSW)RoCKquR
z1Lpj;9ZZV~9FKurzDWg$O&xTCW`<~NY=m#F)bQ_VHzB^haay&-BK%(J3Cgq^)l+Lz
z4a`yGInJs)0}@R=wiX08-%;;{=$+Qc@nsV5rTeMDZ?gBU4AnL#%z)hxsh3?s?Y>vt
z_);j${z__lTrKDYxayl9aibP=W=GIWpw^Q<tMBHWkaR`X2K@=dv$lDztSQi{Go{^(
z#O%D)iSHFfYJx1=r3Qi%EfMGY+&I#J$o*teygv_37cgC*Fyx}G&jpPlaa8U_FH&_a
zkWs~-%er=v;@M3=yLtC0cJ9!#74SKBMTOC-3b@{39*&Z0Hp`kPl{sU$cU5wqVRmLO
zJMm+T^8U8%tp(N9WMvK$;*pBDC`W>rU9_Cn(T|pjSSY_;|EYKXB_3ykz6Gk02}|XL
zWmzOcHlPKiVJljSx_}Syq*Nozv9yeI18{bB((7MFx-cq3{I!A1ilBMtVy91}GN60g
zQbZodBry@XRpycn7J_|2&&Hcg9&dYv5#$LC!^2)xZnNv!4o9k$RoG(6!PlndyUTyo
zxQw$pY#deS0XEJ#I{*2wMv5sdrvNbh{h{6&vnj3Ga%vWr@|@(?71Fqeq1QLO4?K&H
zPu`jB`ef$NOne{8%AxDg7$uB!{EGH0?kNO9YZ;pWv)XaO5Lu1HB`QVX2wj4D_s2IK
zWu2$TuTd*T#aSXbeig`;=CSn^XfF|`amZhfoJ)oQ<ebJNe^ub`VM*<GHeLCDiusc_
zw_fPizl$ft#~4kGIq@T7FN&64C&v|Wl#rEf`{a*3rPXdrwlZyrHtN5Z{m3jv#^4pv
z0}Q+51+bSyX;EVg6)e2H=-7aw=sgJW^YxIK1JeL%*Q>+=L4TBQCzw@*!%e5KTL8Zh
z9W|H}FY~Tdx)(|y9fKHr6DQwtTa2XC-nMUGj9#cmXszLsM!6c(x}KaCkwwEsRfhMN
zYE^qg-%ITflSshml&*`mg&z6m2imLGS~#$|mfo9)>Dt_AAuDK9g3&d7x=+D<{6?GD
zQB?RmnmMB{bh9-b)0xaCz>II{0Q#67w{FF=iXLs`UTf^grp&Pm4b2(E>Qg@f2GANi
zDQW{|@hmgoJ5m9aWZe$b+;*k$U)?BIKJHyn*f5cdE+bYt<r7U=W+C;_h%%U}wlLj)
z#!KR26t-$exhNkp<K{#lz*WUWL&+>&B)vOtBa8N3=ZylFR2x%4To;+KK$8zu+hw0P
z$QePZ#We4H?P&~NgzH1slT{`eRUNAL^aSE=ae4%u<c=VldNwIr<Q{kOfyUU#q<p-X
zc;M+Y%6u3=kQ+>1<?4y|;7_4I@h4dGTT!5R6ZQ-+e7Z`8*1CnnJWvB(&IDQQB12xr
z1S)M0Vcy(&T~%YoFe*O49Tm@`o5a6_is#cUstL1}*T=jwgzuLio(gB7aVwV75KDQX
z)ORmxlnN1M&@6p;u<mrWwfr;ON_>sdD+E{SJ9>Dq)if^@#*Jess?oD<(8P!tz!OUl
z(ZaA~9E;^Nv1J2ys#swSSs_9T3e$yC>Q2X4H^k#u;%mrW0f<rPsgbKhG+6gA2Gc$=
zn)tSjTaITSlTI)^+Qn(dIXQ;jqkF(klHnSgAxlyc86PSJ_5T}yMeE+yPLbnTw&kXu
zHnuvH_9J7kY~W8VFRWoM1Q2>qm>yh2clxn)!(*Ind=2v}1aoRR9bCj}8VZH+;B1R(
zY^)o!G35sE(h@|hFbsvWwVVc9Ht_!Wh7v-bd7W-!Y-nNtE2?$m`<fklERX(136L}T
z;>Cd63w?b7AG2pqgC=6QIOSdms40WZMZ%HR>G<_gsq{=d@bYQ%2c&q&@Jnp4x4SrX
zjxl4f_nKqob7P&=v0g`Z<#fT7v5{t3sq{zqmkG_XNO9BQmj(Y^I(b#yYxXz^?|C+6
z0Wn{#{bl6|F~5b}O4QApJ2D;I4X!-#K-zBj1AV(l=M4$H-{86Np7%HSZJ6f+y}QUe
z8??yYUCzCsmA#?lpNFd*(JJD_F7y7b!{LqS)qY%ivs7sF$@&wWxAIRj^GNNq55<B&
zyd;T^4+dW8;Wa&Aeu&ks1D!Nqh+HS43gL!2BXyrAnUW~)<96$oKU{2j{KWT2?<mr-
zawXK*EN;B=MZf)a!B6)oaag0H^mF;`)28#%LG<bS*W3F=uLYNP2py1x)bSq~%)inH
zi@@ef+Xbo`2ZH04Cw9~uc_Lx0hq4*A<DW|$Hq@x|aO)Fmz{~UM_{h^%@vQSp{|dm(
zId4(?4#3e(0!P3_{GIwzsMwYW1o9>ZfoOl{GItRYq5eTW?*9Hh;&*)!Pb^3M-bhmY
zOod*pVkzR$#H1Hc7kOuq$E=Ptec(+f?X_6kNXJ^6%napbs|X!_^_xsq*-Cgi*ypto
zY$-K0E_<LycKwIY>+VNphIj*-5l$-?m1Nuv5CW+@4AfZO=68_FJE-k&v`UVl_2y>H
z0-tRy(5>H}oD|buzIp_Eo;|wECATNXx?85jPcFvZ7LC&rWN@*p<&RUS5j|)LC<+lL
zce+XWwE9@55|+c!e1vKyj6iQM)%gHia3VQ+f<++n^WEcaF(O*#Rd>a3F7lMGzhKg1
zXHE%tAlQ*14i~da<>nG+W>Q&lWlmeVO-Xsei?%$)RrkkG!kVW`t1xy;!_C6porn@@
zMlKratvFJyxgk>4Zh%PQF-ppP*kYBV0kJ*{nl+V@oihsg5pD@Ks>`2Ij97foua^}d
z9EkTwrWWdsM^8#bpVn>j+no2SFP+d{*o5ux+!5l$&y`hJ{#*OUzb(NILI;RJfO?b~
z%>zlQ*h+@4AUTUyI%FnJb|T_*Wcz#!41E>gMu}LWWM!mVO|BTQJ!qEa#T4q7->}br
z-G?0KF+|t~6~XEri{qCi&2O`oE!jUJZyRgBO_w6vqV-Y6BX=vKFkAJP4l-ok`@Z5?
zjp7ciWSi64D?N{EM$GJAOx}1k0P@{dOG74Ug(cEg9z)Xt9JrSTxqZEDBdW~J>Q=*8
z6^`^O-Ys9Q_(T<O7x|0mpB#fY71;Hv^Zh=G2`LD~`1@Ee9e+P0+z;tu9Tor&cDf)|
zImXoQfh1jQWlY<XE^k&RYo09ZtbC$#mz+-3l1A7g05jJ?!jE!zDNH}_`Z;W0Jxjs;
zATTM)ezmOMN>+E*-sl6hVys^LVzX7;7p7gGnZc6p&YPO*;_2pAm&_Nh6))(N-QXy(
z@ws+y^Dc+!&PxMVOAm<`Rku>L`@Ejm_J)M*Oh_0-@-odSc~wc#vx?H~K?^E>QY#*`
zU*#6~LknoINm#(#OeK`*Dm!$!&n)4t=s0mu+l-VirxD*Iws9lT#Iw9uch5BW8S8dm
zeRzdd{tr!Czdzy!5(JvAhCVIY#`AG#@ybN>QKHIkeWLwQCsMKn-(b1cG<`g}a$WJ(
zDf3t)8#$&&T4OPf<j!vAXGk6o;rw35X?+0hSjL5sMgw)AU|~!|%tZ1xPtes+)&Kj+
zfIfgKTwG*v5seG*7tx^kcRr176H&aFDG-GaYciobML8~dbD|L76$Auw34$YnCB!e<
ze}CkEUBYhz{#l}k3OWDBLHxJyznT2EC<qiqByjalVV59us2C6T#a{oLCV#JvA_8-x
Uo5VP2z#uRP`0bdSc`l~@15@!G<NyEw

delta 7016
zcma)B1yoe+)*d>fq=cb{9y+8MI%F6c1VNAxL<EVU1w53rAT22|ARtJ0E8Q`4mr{b#
zAjps3UEjU;`|e$L{b#MS-gEYT_FnJVXTQ&W<M7_vLn0jn7B(da1i}aHD#a4D<L@fb
zo0V9b;>S*K0_T~<-$uAA#^}$at0}|rxgEP^YHqt&&8XTfF><1~sODpRf9&qoR*pUC
z!=Etmtte60;c+v<GZWW>RziICUxS>V#h5fmNTlk_*BBz0T0HPni%9F0mJ@~)a+C9p
z$X{4kbFlPQ>JixJUQ9-x>1lc$+U;N%a<_p+#TGFl0opgd-q&?6jcuJB_jMUdJeDw%
zD4cYzgGs0$xrfU|y%X+c^Xx$a4s8~dCbFdcs`uBL94JHhjR*jdI>`Ip(%x@oKje{G
z-sGdk9bWBkE$%98J$!e}vLspN*#GR|mFKFr@_5*L98Hx^*sZ@}&nN|lKk1^!&L^Pr
z^oO1^A=Jt3`+ZqvnL594{yj>SWSIUqn7c}qad>JNB){WTV9o|RCjx;!SjTVTGXkAv
z#;$WU)Gzl-M+#hjB5irCSTQ<JG@3dQO>htWQ_XU_hx&XUmBo9nj_-glO!_h8;m#N2
z@Jlyp|D|((rO|<t+8^tsTdtXRUhjWyf7&qa%O8>K@j<j17z%CPnml7?g0H*4Xio^w
zT@KKxEgA=lL1JJM6AAd(WZ4vxDBxLpT>b&R_VV(L`BBKZR^^9i2*GRfNChAP;d~)Q
zf5IVma?pM@r$r(9EkNsft6m4sq;XE5`F&H`omv4Sit|F7%RNWAqtNrGvrIAI_)C}f
z^`Wy&r)260;lGCVuP*1@zWC*C`hEKnblK9w<i;yHU%K|goDdcpN}gE#D-b}=#tU4p
z|2(%S8@~)B=Xd@25KoV37JwA9R&$?L?mwSqmC;A0oBNFW0<g0&0GesDi<<H3Ve%;r
zTK$$)5bfTz6zzH!!vLqAi8TLWsjK#&(ThjZF3xJwVZK!>^2dl&5}aBNx1LZT^|d8{
ziXrw{8Z;d`vv^ogL+aqW4B#-KA4f*<3DsD)i_H~Lq=4-ktj*+np7@sGF!{bR)_5J`
z){b{T90Xn)ps|hY>Ltuw?h1e47B|VPU}gM<_iZGfqO;-~aXe`<gIK#*%_c4KB-N*}
z_j0rx5NuUFYz?1ySr7$972x5k(9kB{1n`zYH17v8S7}9edgEu=?0{p5h#*3^szaE3
z-CMLf$@*ZBeAxawN?rAV9Ku0`hmxVDxHGa?#r+BO!<to<njE*p(Pi~t)ZRQ-#gLq&
zHm}&Wu+r5oVH$yFtIu5R&m-^HlqazmIP&`ZVDVW{NicbcFfQhr<4%@rc4j<^Slm0Y
zzAO<MO?t-17JmGoBpMjANr{RoZmGA8>AB0{QcY~PK(|}daYjED$6O>)7saa0ee2$*
zJn<U(r%S{Pb=gYdZtEek_1TZb-Bv@i7Di~r;z1Vg7M^@%th`k_lSkKpRGM-uczpOE
zq`_4OK#9FyuwrQ)OAhAs0DdN-Xv#@juOiBqQ(kDYcdzrT1iJ#~slst6rkaPZj*1+(
zr^LHNG4btk^BvXW?nB}832sVwj+-OI4E5RCj+@iO8FktE;%=(21h=#dp8@{)C&`YR
zeZ*#s*(5Idalz1V{ZD$Dox#hAzfIXJ)yFQLX$;;ThCus*)iY6Hz?n*mW$?Blf)Zri
z>Xfb*qn*A9fK?0xdpZ>|n%2n+{Kyp1sQfUU;}m9f$h**$Dsnb-k6N51cD;l!|J$z#
z@3HOf1arKJ)D~4uf*_K|j8w~{u$0*fN2=OWWhGV!gx<ANxX^1%38eb>$Aj-Kv&t+V
zh@`XU+odod+Tb#k@u?G;2Jm?0arQhlCT+9^s6`G4z2`uvn91&!DM-O3{*^DtG;>NV
zS<laLgBNL-_x`+-YyFUAfT!NpI#t}%S;l%m=)(by?<4TLJrROIasz5}_2dG=x+Y<M
zZ@*+Ns}~%>WM^0ltJfIwkERNWq6maslhjgg`_<^>#5{)Ui`*S&%O)_-o0Ss78=l4#
z5=!m^a$Q-&C&josi6bV6yW>Qa`YNUShs@o`P$@AwsRKj?hWLsfh|co6P;=ER?8b(A
z#3^qtsBR?%-1oHiyQQsi%k)d{Q#~y!NpH2Gy};r6jo8F~){jZu2jJy_YX4Wq6()Qs
z#z_GSZ7;(uEqY}I$wq$AgTp{S#j$iPuAdx&fItSiN;g_5Ast+Y<$12m84J+KjFtSa
zuhdw$&bn6EJ=bIM@>M)Lw`V^Y6VZ5f*I^GdOcK3~r>mVQ5nbz_-lwMMpYD$7`U(^4
z9&&r`k-uR}R9@Ku1u@XzOWX-{HMAh^t9~>gYDtT;bJoLf7C(5$s$2-m%YAj|6iu8l
z16Ytn^5K?GTm5cCS*Wqjyllp>@sZy`q$YE1$Nk?f_Dc+_PF;Y!eAY<(><2-+o<X;l
ziGUv#fLEC`O&U+ULwR?h8OIEmFUB<FI4%XBM9kr*dwNmk*E{Cp`rW-nn%Xtq<MGcI
zkI*#6Rr#E6_qH{)y>NLSlOKW|UVH+Kanl&33-?Elg6}mU$0@X?Nb2X?y~i5y5ABYh
za7HfeTJMkVq6_jpoje>!<Z@>Sx-SR@wO+%`8)wQ})<mQ?8q&QdZ6Yz~^5kG1o`Pbv
z{D;-GrSs#aE}pKP3xiai;zsierZcdt=Uo4Itk;KCwr4#r?EK>f{uCR>=(ySl71Xcp
zed&9vu;MuXq*e@E2ERIs#iN6ki($QbA3(l`t6oP}Ao<}K4E0>r7Bdek{Inone<<Uw
zWGhC~VGyaHQT<0loY0x|8wwRimotjnWkr!<MjdrATug@GmI+Me-7`q!c=qWaQL;qi
z7|<f=@mdPxaz>K)E@zX%isUzIySbS+r@9G)DJQ?4R8?SOpA}<bz|VIpVW-BA&Ugj<
zmJ>;vKFSY!SnP?{U1JjWtS+*4>_$6zG*aRBjA-I0${86#vgpit@gG5){SRMRI>NGh
z=SugF5ufSR+Q$STs4P6$?`4?nV`Y#B0GP3VjZj18AQ8X$hX($b8Fr3MkbNnb28m6V
zOkvLPo(exH6-OrRSn^<~B@Vj@^T&+DAn3kCsgMBj=;(Q@*}R6*Ri4vjG8KNoc_I$2
z{FXh_^p5~DrZq?3<&q5P;b?*t_gW3xo_XfaS<eG7B(rXPmHm7xW+e}=@xWq=5h#8v
zq?5ZVBk#YGfq{{9xfLS27chOv#k`XlCvCm=1f%Es`*z=et&4y{PC;VaM?Y=rr8{5&
z#)98xTkGbt(G(KIxI>eq>E4*O?QQnrTb^M)n1>e2)jpNdY_I?osG!y)Nfd>rMbJ5B
z?lBgtwgn|o6KZQHt!+>6M43>X3aHV0Wj~MM9V)XO9JGyU=t=Q8o65LJhLtRxLPt?W
z62`$#3WIuPF2VG$gI7TZc^8}bl3oc!c!?TUZ)lnEVhUEW3JUXOZ-9QE75wyN#u_A-
zx$#zGrN|Sc+}m9ay&;SftO^tQ$zay?<xwSaw37RC=nXo9O(BC#*A0mCa*M-uk5e9n
zua3gW(QtAV&hUGja2-xKQN}L$(!Ov9YYrPUhb<(`gW!~XQ}+S6!2@z7%SP;$k!R#b
zJpu%M0t@mlm!7K193X-Y5V<`9$2|hXeFCppTG6bnBG6Wmh)cbg%dxD>G3Wu}KS02r
zs6Mg87nDi3R7t8c`Fb&7V_9Ki&^d$cQ8*Y4r&i%ixyNa&!)Z*JVa=Fft&m}@>lWwj
z7Kh^=r!orvG76{u|AS}?;tvCoI-T>DCco}a14}v+UoxIUwGW5H&?(-wbc9_^mBc*2
z;!FrTY~SItNxEl^rS72gT$$7jtd8MrE3MoU9BUySNTe}_@c?0um7zR!C1%B2sy8ae
zhQgJ_UX#(heY3GYoU@>GTN>C{KcI~^nZZ~pGvuz6B$PHw{VozZ34?r`EAqM?W$7Wn
zjV2#SeWMZR8^5vXYx~^SmJ7uq<-v-{A%YbXu|x=ml?jJI`yikCAUV(=e4e1N93qzg
zp|HB5vs8XS(sAM07*>Y6*{W<Z-dsa`VGe9pKOecC^<-GIo6EW;cFab6IxYd2eClhi
zosD<ynJ50Lb$(4}e{>ooJKh<|ARD@(J^>6C4%j-iozxRep)T&W*6xRX(72kIpF6cj
z6RR(_#5_X>G`GUfx{Ok*o@7Qz3~*{+%yf)*U7fk3T1xSC82FjOeWP@KqI?Z@Fp+Gr
z+K{o>9^$Zc#5$(($x|NzoM?b47v-lX>*6u0$a0TFIRr1-91$7)8MB&nrK#d1^Cie`
z=XuNe3~dbK#rEytW(8w?72wF4Y0o#@<tO2q>YW{FBSz`cpeEf*uU^m8)3HnIS%)T#
zi#2%bS+*nS4H39?J^qnU66c{^+-Za_@^GF`l#X<XNt>~+!eF@nB?EBe5&td8k0&L|
zYzjYiFob4(d@KE$ZM_wTL|o)n#p0Y%(dK7@qDaU-3~xJUQ2SkwJ^$>KIwI(N3un#m
zy?m*+96<P>F-X@7bF(oS<4U|}89eThd<e$YseRmf7Xw}LMa~abDv*1$L>KF5L1lS~
zrX|7uBV=W?(o+C*;sj92Cs#LCZ!ZmNRxj{%32?@U$4|imsp}%XA(3aN_D(vTojhG;
zPSMukZuTV!`>kqCq$=p^R+q&z*zvvbE(nm|GiZ&daM9b-KlYLAzsOi`pK@XLEI%1{
zUG(YU+>U2!U-L4ZCUv?l?cPBseR&rXkxAG*?n|p30WUT00b-BWpEg`wZ+wxHd3yg^
zu6OSm<KNd$^k!i#FuYM55Qu~k1Y)^adwMwt*m$|wARR3oH0&JhUH|eX#P9XoAtMFh
zL?=$>e;twSleaP8oNV8LtB1nNRUtj*Y9Kgi5Gx90E_96M!W8ji12&nTlZo)ngJg~p
zaP0!VjCQ{(PdG7mGa|nyO=64a<ot#VxlBU5Tu(KMpc}V>?H(^l)4r^7e>}pZ<zdo2
zzP07?{;s41KW;PJ$IXPOKQ4FFYa}fxd5u&$e%xhLI-UMbr`yQBd>kr6Vc&Xswt02O
z$4A6uqE;rl(8bNo4M6Ic(lPkLpdsTniEBAaX_-+W@oVwpaC`eEKV;oFc%LD)B12_v
z{}i_qb>g#AJhf!^%_w<fyU7qlFtU5@bZ59F&F>s)>X_Dd-F$VtJ9HI!M{ba5{#M@H
z`RUg~CdcL%XAm8!Mn6}N!<h5W?@J$D%J`faK~^XF4*^CwAa&{dXw|0R!lqqVjW5&h
z*UauriSt$DOQCG^JEp-_DWfs~kn;IuZ+v(;$9dQolg1)DJqMh=D7X;V8{C&GB$O&`
zuC{Nsb83A%`LTw&^`uWc!`bYhMRzX<$<)6h%;;BfKUyii?qD@*^Yp_9NB4Ts1v}0B
zwS$wHy=?|QAoEE5Yccjm#VG=#(nmv3rk|_Bjg^gZL%?}{<|r3kadqmHNpnjRY%ULN
z&4JdKx`&P639}KZia*Mzm3NwA%US5$s4yd}ZH!Qx_0Ht<zS3qkX1AcrPU%o1|IB8{
zY^?g$BaV$9*ZThBxrFV4tCT0aZyvNgC=A4jvI%6S0{V)Q#PZUzvXm86dnqYRbEjY2
z1y2eu)-Z&L@1&NiS>1jti4!@gq$Wb?8x;}x7U4E^O$6nsFaC<`OE6I}Fl7tz_%6x)
z(I(LTF~<du(-(DiqdwCt=e&W9)%<y3W~Z+?jpTC!iSUn6^*SqklUFu5i@Dl@Ps=en
z;AGv9R3M%{D2oD18S4i@2%Gn!JUj98=@G3%{UHNLzV~-7Ztiy$yLlLS`T<AcWvv;)
z=Z3dmzX)6>mPfS&9DfVUt=eJbCPUmtnsg@Xa8259Q1C0jy|MbGwJD)VweY~wYRmvC
ze7294QUO(>WhJgEa&0wYG!x95l6qudcVeBf383xcP@+Op1T-7;IEn1LBVHP+Tt^Sx
zmNS&d?m?wXoI(FMt7%&3ZbQM1X9U(mDTmsRF%Oz#H3yb8qs6Flo`FB>^|zUj2q4Uu
zbOUro=BdKOg)sDXiIPolCy`zZ+ZSP-4L0a<o%aU>HR<2V4fOh#tBEYnJ^ebY<_bt?
z04J*oO?~=2U!j$nL@Em7maLlBk6IGG7_ZYPaG8q`<rKetIyo-&1UEZbz=ql?8|8bO
zM%{(@qvy~{i4G3tyflF_$?XIVRco?NfjY7Ba3LX<m^o92da#%zR5PkWr_VuA%q5Tm
zw=eB2GQLw0PLV4vL}Sz3F`H!UgwnEE0{qpBX!P^-4E5HDD=Byi8&Iy8lJ(t9*ev&>
zXoH|16%1W43&kGn(2Md#F51PtK?_TTjImWsBDwS5)G(}L_d#aW^!j^DNSve^Wphw{
z9qIa>{ptQ<6y_?#ky3C=id%EgNU0xywD644tlBCzx12Co9#t0p>Un}nMw9;`fDT_M
z6&sacuJ$f2hlA6mr&V3Zz3IMgDAL|dm?X&SeO!oY2|p%|Na)#_wxK*;QL$gL-G1gp
zpuU~<U@2=r>Ojf^YG}$3Ar=~U!7(WgodZ?O>$|)wuoe%EJK?DLEl{WB4mZ(<F3_u|
zrcbYDZH8Bf8lbO3pmR3@;J0L^H*)577rgyB3A>6GFB1L4yk+a{YVd&qs~l|!WxY*o
zZUuW8fwtT^Uru;BMMX$PULv@**u_8$d-8i_ek!YEez{HIv;HF0PPxf6aaS#ICx##j
zA`#v~!`YdN3|*IK8|h@Z9ms6bI|{vI!h;aL+{70(#kNE+^-w@NI51QRMGrRKcv&<*
zobXet{+)g3!|lRkk6oK8SAn3NyW+C8J806U_Q@0H`+LA&i-Xy;tL5fRQBcJIf%yJg
zaWF^PIQ*9q!T($QdSqA~_$*I```w7(Cn$v`TMdpZm!T}8DqKvqj^mvHcB9GO$$w|V
zH`$zc1oQrmD%Bs84e`+_SZ-e%z-v8!yGzc_X<6m;>R7iFbs}jWzDv`V%*Va=__1y%
zN=(azFxS5Apkj?!F_n$gITV+tLg}u_qRC~<(`}Ok389VGikd(h;4hOsZaVa8k6)si
z*918Q*(jza4<dR0xb9#RHt_yy<PMI;h^IBLgWP`T*TJ%vOqq+zY_cI&3p>9oE3dq0
z4PzRQxi!ah=}%k)nlCJeHK-!ATVpDo8g2{>7O~>0;I=Lr`u?q=j^ZnO$>M`RA+JH8
zN)t+B{G0vB_-gu7;P>|A+@flb%S|*vSRfGTKO*YsgtWJEa<czjTTArXwLS|HT_tie
z$s^eFg-ROz{o(jz8gr6}`3aWr4;>H1TQWuR{IT&!(&^<mPx^f_J=e`^l}l$v!FRYv
z1#2r|q(vA!pJ}U?t@1dW*t^mD2O}@uNQr<PG=WFdoouz_84m`!A8E_#qgE753Ob6~
z%g&);zC!zJ;OtxGr$0rcE!z0N>-y|tGyXliQdMWT4dkzS6qObjynR`_$C{wpT?VVZ
z(Bk&HMNHoj!>;VL%*@@yQkrJl9^m14mM3QFVg>c#J#H7T7BKd}AC)ERp{zZ~pS-kN
zYk=8nS!Zxus(PoFzl2lIiH7)7n#*eIy5lOR(Qhv}bi*#w$l-%RB!&Z*<^x#`syI`q
z6KZ-=M=wDcg*#w}pdQl-W?hZcjyiOZEWNRx{@k!2^`#!^rK+BD6qEB9Oo^B=J#eWE
z;JAY4X&AiB&B|8AYgnov+gIMX-|8iV64*D(yt83*Wa~6=zk_i>1}>v6=eW4a@Yj^j
zP<8TdzxgpqnDH`9%s`wBNsu7XtI|NEKltfHHfBD*S_U^iS-|;i=5Q;cS98gqK(SJ8
zmJylXnofI>2g>aAcT;DF^aM%N^|?)+_+VvKbSjMXK0v#^7jt|ALrnQ6+6d4EOAwVV
zpD?l@PBImQj4!)K(e6g83kjtdqSt~UwJtK%VZ4Eg4Y-n2f)9#HG|FJ3b{^CUkMcfk
z+-F-BgwWGwByEpwE!xg{UP`{>mwglMrKd(WRn&ZT81Msj-O1+{Yd>n<zR?d_q<zBl
z*RNyE1=ow+q~A(BGlU)(cKjqr)McKxO;v3vy(@^3{Jhp~t?W*ZUb5<-TxUr8+^`+S
z>u~ohq+P=7ub0uSRh2Pjpn;l6iS)0;x?Yr+_puDEwy-3k(XuA+rq?{TQO`CAJFeU=
zAPJt>(S+*YbE^}gYE!ztM>Ut=YR^MhU^~obn}WhVv!5)=FiryC7yEdj!gUfH$J0pf
z;MPXKxxUKFtnv$YcV|(YEcYYiVV|1Q7&2)Z98h)Ff1|7;h0)K<9w*VrzMq{v1i#`F
zu75F?&uFh;ir1$Rxqa0`=@e#&Eq^X)Pqa!Mqh}f0{W}4#AaRVctFb*}xx;w9-w2EB
zXP&)Y|Mt{Zz?C^&x%uj#{2M=3gd?rf_bnXh?d$P-;D>HwP0yxQd-o<8uFx$O67Lpk
zYkzHynV1wV1PC<9&8^buATTf~vHm=#iFaUz{qvwEzKmJ?_h{c@miygsv&jAZOu~~z
z;twA9`?UKe<Nmn(F|MipKa=pkc7opQ?@awKe`NH3^T+>tR{y_<`>(#^%US3#l^EiO
zS?GR$(=H1=P|OYj{lmc@ME_U(e{$Tv`jW@-M_>M(@&0c7uki;0f&a9;Y5g&jpnn*f
tBjeSWsN&OEi6FlT^`9Ay4FXa81rOtj?_p&iAp#MBxNq1tm;1N&{{Y4b-!}jN

diff --git a/app/services/publish_export.py b/app/services/publish_export.py
index 7dbbd17..03238ba 100644
--- a/app/services/publish_export.py
+++ b/app/services/publish_export.py
@@ -4,7 +4,10 @@
 ====================================================
 公司流程: 每日把持仓与净值导出为 excel 公示。版式**不重画**: 仓库里放着由公司模板
 (量化数据2026.8.3.xlsx) 裁出来的空模板 app/assets/publish_template.xlsx —— 字体、
-颜色、边框、列宽、数字格式全部原样保留, 导出时只做三件事: 插行、填数、重算合并区。
+颜色、边框、列宽、数字格式全部原样保留, 导出时做四件事: 插行、填数、按最终行位重排
+合并区 / 行高 / 红绿条件格式、画净值+仓位双系列折线图。openpyxl 的插行既不搬合并区
+也不搬 row_dimensions, 这些全按最终行位重建; 模板行位之外的行属性进场即清 (公司原表
+把老平仓行折叠隐藏, 残留 hidden 标记曾把整段合计行与图藏掉, 2026-08-28 正式机实锅)。
 数据只含**系统接管之后**的账本 (2026-08-28 拍板: 不回填历史, 启用日之前的净值与
 明细继续查人工表格)。
 
@@ -248,10 +251,11 @@ def _nav_series(snapshot: dict) -> list:
 
 # ---------------------------------------------------------------- 版式 (模板填充)
 def _copy_row_style(ws, src_row: int, dst_row: int, cols=range(2, 22)):
+    """只拷单元格样式。行高不在这拷 —— openpyxl 插行不搬 row_dimensions, 拷邻行行高
+    等于拷一个错位的旧值, 行高在 build_workbook 里按最终行位统一重排。"""
     for col in cols:
         ws.cell(row=dst_row, column=col)._style = copy(
             ws.cell(row=src_row, column=col)._style)
-    ws.row_dimensions[dst_row].height = ws.row_dimensions[src_row].height
 
 
 def _put(ws, row, col, value, fmt=None):
@@ -289,12 +293,26 @@ def build_workbook(s: dict, nav_rows: list) -> bytes:
     """把快照填进公司模板。纯函数 (不碰库), 可离线测试。"""
     import openpyxl
     from openpyxl.chart import LineChart, Reference
-    from openpyxl.styles import Font
+    from openpyxl.chart.marker import Marker
+    from openpyxl.formatting.formatting import ConditionalFormattingList
+    from openpyxl.formatting.rule import CellIsRule
+    from openpyxl.styles import Font, PatternFill
     from openpyxl.utils import get_column_letter
 
     wb = openpyxl.load_workbook(TEMPLATE_PATH)
     ws = wb["Sheet1"]
 
+    # 进场先做模板卫生 (2026-08-28 正式机实锅): 公司原表把老平仓行折叠隐藏 (24~97 行
+    # hidden), 裁模板时 openpyxl 的 delete_rows 不搬 row_dimensions, 这批隐藏标记曾
+    # 原样留在模板里; 表体一长到 24 行, 卖出小计/存量合计/底行/图表数据整段被隐藏。
+    # 模板已重裁干净, 这里再兜一层底: 模板行位之外的行属性一律不留, 隐藏标记一律清掉;
+    # 残留的条件格式区间 (原表手工长出来的死区间) 同样清空, 后面按最终行位重建。
+    for r0 in [r for r in ws.row_dimensions if r > _R_SUM2]:
+        del ws.row_dimensions[r0]
+    for r0 in list(ws.row_dimensions):
+        ws.row_dimensions[r0].hidden = False
+    ws.conditional_formatting = ConditionalFormattingList()
+
     n_h, n_c = max(1, len(s["holdings"])), max(1, len(s["closed"]))
 
     # openpyxl 的插行不搬合并区 —— 先全拆, 插完行按最终行位重新合并
@@ -360,7 +378,27 @@ def build_workbook(s: dict, nav_rows: list) -> bytes:
     for r0 in list(range(hold_top, hsub + 1)) + list(range(close_top, csub + 1)):
         hts.setdefault(r0, 20.0)
     for r0, h0 in hts.items():
-        ws.row_dimensions[r0].height = h0
+        rd = ws.row_dimensions[r0]
+        rd.height = h0
+        rd.hidden = False
+
+    # ---- 红涨绿跌条件格式 (原表样式按最终行位重建: Q/R/S 红涨绿跌字色,
+    #      持仓 S 列独一份用红绿底色, 0 也算红; 颜色与原表 dxf 逐字节同) ----
+    def _font_rule(op, color):
+        return CellIsRule(operator=op, formula=["0"], font=Font(color=color))
+
+    for rng in (f"Q{hold_top}:R{hsub - 1}", f"Q{close_top}:S{csub - 1}",
+                f"S{hsub} S{csub}"):
+        ws.conditional_formatting.add(rng, _font_rule("greaterThan", "FFFF0000"))
+        ws.conditional_formatting.add(rng, _font_rule("lessThan", "FF00B050"))
+    ws.conditional_formatting.add(
+        f"S{hold_top}:S{hsub - 1}",
+        CellIsRule(operator="lessThan", formula=["0"],
+                   fill=PatternFill(fill_type="solid", bgColor="FF00B050")))
+    ws.conditional_formatting.add(
+        f"S{hold_top}:S{hsub - 1}",
+        CellIsRule(operator="greaterThanOrEqual", formula=["0"],
+                   fill=PatternFill(fill_type="solid", bgColor="FFFF0000")))
 
     # ---- 表尾注 (模板之外的诚实行: 缺价与未摊费用不许静默) ----
     note_row = sum2 + 1
@@ -376,7 +414,8 @@ def build_workbook(s: dict, nav_rows: list) -> bytes:
         c.font = Font(name="宋体", size=10)
         note_row += 1
 
-    # ---- 净值序列 + 折线图 (图占位与标题同模板: 汇总行下方, 横贯 B..U) ----
+    # ---- 净值序列 + 折线图 (双系列同原表: 净值主轴 + 仓位占比副轴, 标题同原图;
+    #      每个数据点带标记, 首日只有一个点也看得见; plotVisOnly=0 隐藏单元格照画) ----
     base = note_row + 1
     NC = 25                                # Y 列起三列数据, 在图区右侧, 不碰表体
     hd = Font(name="宋体", size=10, bold=True)
@@ -389,14 +428,30 @@ def build_workbook(s: dict, nav_rows: list) -> bytes:
         pr = row.get("pos_ratio")
         _put(ws, base + i, NC + 2, float(pr) if pr is not None else None, fmt="0.00%")
     if nav_rows:
-        chart = LineChart()
-        chart.title = "量化净值数据及持仓仓位变化"
-        chart.height, chart.width = 9.0, 30.0          # 与模板图区 (B..U 宽 × 16 行) 相当
-        data = Reference(ws, min_col=NC + 1, min_row=base, max_row=base + len(nav_rows))
-        cats = Reference(ws, min_col=NC, min_row=base + 1, max_row=base + len(nav_rows))
-        chart.add_data(data, titles_from_data=True)
-        chart.set_categories(cats)
-        ws.add_chart(chart, f"B{base}")
+        last = base + len(nav_rows)
+        cats = Reference(ws, min_col=NC, min_row=base + 1, max_row=last)
+        nav_c = LineChart()
+        nav_c.title = "量化净值数据及持仓仓位变化"
+        nav_c.height, nav_c.width = 9.0, 30.0          # 与模板图区 (B..U 宽 × 16 行) 相当
+        nav_c.add_data(Reference(ws, min_col=NC + 1, min_row=base, max_row=last),
+                       titles_from_data=True)
+        nav_c.set_categories(cats)
+        nav_c.y_axis.number_format = "0.000"
+        pos_c = LineChart()
+        pos_c.add_data(Reference(ws, min_col=NC + 2, min_row=base, max_row=last),
+                       titles_from_data=True)
+        pos_c.set_categories(cats)
+        pos_c.y_axis.axId = 200                        # 仓位占比走副轴 (原图同)
+        pos_c.y_axis.number_format = "0%"
+        pos_c.y_axis.majorGridlines = None
+        pos_c.y_axis.crosses = "max"                   # 副轴靠右, 净值主轴留在左 (原图同)
+        for c0, sym in ((nav_c, "circle"), (pos_c, "diamond")):
+            for ser in c0.series:
+                ser.smooth = False
+                ser.marker = Marker(symbol=sym, size=7)
+        nav_c += pos_c
+        nav_c.visible_cells_only = False
+        ws.add_chart(nav_c, f"B{base}")
     for col, w in {25: 11, 26: 10, 27: 10}.items():
         ws.column_dimensions[get_column_letter(col)].width = w
 
diff --git a/scripts/test_wiring.py b/scripts/test_wiring.py
index 06df5cf..8e4d34e 100644
--- a/scripts/test_wiring.py
+++ b/scripts/test_wiring.py
@@ -2494,6 +2494,44 @@ def _():
     rows = pub._nav_series(s)
     assert rows[-1]["ymd"] == ymd and abs(rows[-1]["nav"] - s["nav"]) < 1e-9, rows[-1]
 
+    # --- 大表回归: 表体越过模板隐藏带 (2026-08-28 正式机实锅) ---
+    # 公司原表把 24~97 行折叠隐藏 (老平仓行), 裁模板时 openpyxl 不搬 row_dimensions,
+    # 隐藏标记曾原样留在模板里: 7 持仓 + 10 平仓时, 卖出小计/存量合计/底行/图表数据
+    # 全落进 24 行起的隐藏带, 导出的表"最底部的合计行没了、图也画不出来"。钉死:
+    # 行一律不许隐藏; 三行合计都在; 红绿条件格式按最终行位重建; 图表净值+仓位双系列
+    # 带数据点标记 (首日单点也看得见) 且 plotVisOnly=0。
+    import re as _re
+    import zipfile as _zip
+    mkh = [dict(account="甲", code=f"{600000 + i}", name=f"测{i}", amount=10000.0,
+                open_date=d0, upd_date=date.today(), structure="平层", cost=10.0,
+                qty=1000, price=10.5, days=10) for i in range(7)]
+    mkc = [dict(account="甲", code=f"{1 + i:06d}", name=f"平{i}", amount=8000.0,
+                open_date=d0, close_date=d1, structure="平层", cost=8.0, qty=1000,
+                settle=8.2, days=6, pnl=200.0 - 40.0 * i) for i in range(10)]
+    s2 = dict(s, holdings=mkh, closed=mkc, price_missing=[], fee_resid=0.0)
+    blob2 = pub.build_workbook(s2, [{"ymd": ymd, "nav": 1.001, "pos_ratio": 0.35}])
+    wb2 = openpyxl.load_workbook(_io.BytesIO(blob2))
+    ws2 = wb2["Sheet1"]
+    # 行位: 持仓 5..11, 小计 12, 卖出头 13, 明细 14..23, 小计 24, 存量 25, 底行 26
+    assert ws2.cell(row=24, column=7).value == "=SUM(G14:G23)", ws2.cell(row=24, column=7).value
+    assert ws2.cell(row=25, column=2).value == "存量合计", ws2.cell(row=25, column=2).value
+    assert ws2.cell(row=25, column=19).value == "=S12+S24"
+    assert ws2.cell(row=26, column=2).value == "初始规模+已回收益后可开仓总金额"
+    assert ws2.cell(row=26, column=17).value == "=U25"
+    assert not [r0 for r0, dd in ws2.row_dimensions.items() if dd.hidden], "行不许隐藏"
+    assert max(ws2.row_dimensions) <= 26, max(ws2.row_dimensions)
+    assert ws2.row_dimensions[25].height == 18.0 and ws2.row_dimensions[26].height == 40.5
+    cfsq = sorted(str(cf.sqref) for cf in ws2.conditional_formatting)
+    assert cfsq == sorted(["Q5:R11", "S5:S11", "Q14:S23", "S12 S24"]), cfsq
+    zf2 = _zip.ZipFile(_io.BytesIO(blob2))
+    sx2 = zf2.read("xl/worksheets/sheet1.xml").decode("utf-8")
+    assert not _re.search('<row [^>]*hidden="1"', sx2), "导出的行带着隐藏标记"
+    cx2 = zf2.read("xl/charts/chart1.xml").decode("utf-8")
+    assert cx2.count("<ser>") == 2, cx2.count("<ser>")   # 净值 + 仓位占比双系列
+    assert 'symbol val="circle"' in cx2 and 'symbol val="diamond"' in cx2
+    assert '<plotVisOnly val="0"' in cx2, "隐藏单元格也必须照画"
+    assert cx2.count("<valAx>") == 2                      # 仓位占比走副轴
+
 
 def main():
     # 静音日志: 本套里有好几条用例**故意**触发异常与告警来验证「守成」行为